Alvarez-MacHain v. U.S.

Full title: Humberto ALVAREZ-MacHAIN, Plaintiff-Appellant, v. UNITED STATES of…

Court: United States Court of Appeals, Ninth Circuit

Date published: Sep 11, 2001

331 F.3d 604 (9th Cir. 2001)

Opinion

Nos. 99-56762, 99-56880.

Argued and Submitted June 11, 2001.

Panel Opinion Filed September 11, 2001. Rehearing En Banc Granted March 20, 2002. Argued and Submitted En Banc June 18, 2002. Filed June 3, 2003.

Paul L. Hoffman, ACLU Foundation of Southern California, argued the cause for appellee/cross-appellant Humberto Alvarez-Machain. Ralph G. Steinhardt, Joan Fitzpatrick, Alan Castillo, Dilan Esper, Robin S. Toma, Erwin Chemerinsky, ACLU Foundation of Southern California, Los Angeles, CA; Alan Rubin, Epstein, Adelson Rubin, Los Angeles, CA; Peter Schey, Center for Constitutional Rights and Constitutional Law, Los Angeles, CA; and Thomas Nanney, Morrison Hecker, Kansas City, MO, were on the briefs.

Carter G. Phillips, Sidley Austin Brown Wood, Washington, DC, argued the case for appellant/cross-appellee Jose Francisco Sosa. Charles S. Leeper, Glenn S. Greene, Spriggs Hollingsworth, Washington, DC, and Lee W. Cotugno, Kalisch, Cotugno Rust, Beverly Hills, CA, were on the briefs.

Robert M. Loeb, Attorney, Appellate Staff, U.S. Department of Justice, Washington, DC, argued the cause for cross-appellee United States of America. Robert S. McCallum, Assistant Attorney General, John S. Gordon, U.S. Attorney, Barbara L. Herwig, August E. Flentje, Attorneys, Appellate Staff, U.S. Department of Justice, Washington, DC, were on the brief.

Charles J. Cooper, Hamish P.M. Hume, Cooper Kirk PLLC, Washington, DC, were on the brief for cross-appellees Jack Lawn, Peter Gruden, William Waters, and Hector Berellez.

William J. Aceves, San Diego, CA, and Jennifer Green, Center for Constitutional Rights, New York, New York, were on the brief for amici curiae International Human Rights Organizations and International Law Scholars.

Appeal from the United States District Court for the Central District of California; Stephen V. Wilson, District Judge, Presiding. D.C. No. CV-93-04072-SVW-06.

Before SCHROEDER, Chief Judge, GOODWIN, O'SCANNLAIN, RYMER, KLEINFELD, THOMAS, McKEOWN, FISHER, GOULD, PAEZ and TALLMAN, Circuit Judges.

Opinion by Judge McKEOWN; Concurrence by Judge FISHER; Dissent by Judge O'SCANNLAIN; Dissent by Judge GOULD.

OPINION

McKEOWN, Circuit Judge.

We must decide whether the forcible, transborder abduction of a Mexican national, Humberto Alvarez-Machain ("Alvarez"), by Mexican civilians at the behest of the Drug Enforcement Administration (the "DEA") gives rise to a civil claim under United States law. In an earlier, related proceeding, the Supreme Court acknowledged, without deciding, that Alvarez "may be correct" in asserting that his abduction was "shocking" and "in violation of general international law principles." United States v. Alvarez-Machain, 504 U.S. 655, 669, 112 S.Ct. 2188, 119 L.Ed.2d 441 (1992). We now address the question left unanswered — whether there was a "violation of the law of nations," a predicate to federal court jurisdiction under the Alien Tort Claims Act ("ATCA"), 28 U.S.C. § 1350. We also consider whether the Federal Tort Claims Act ("FTCA"), 28 U.S.C. §§ 1346(b)(1), 2671-2680, provides a remedy for this cross-border abduction.

In 1990, Mexican citizens acting on behalf of the DEA kidnapped Alvarez from his office in Mexico for his alleged involvement in the kidnapping and murder of an American DEA agent in Mexico. The arrest of Alvarez took place without an extradition request by the United States, without the involvement of the Mexican judiciary or law enforcement, and under protest by Mexico. Alvarez was brought to the United States, stood trial on criminal charges, and was acquitted. He then sued his former captors, the United States, and the DEA agents, asserting a panoply of common law and constitutional torts arising from his abduction.

This case, which has been litigated in one form or another for more than a decade, involves important issues of international law and sovereignty. It also implicates our country's relations with Mexico, our neighbor to the South and an important ally and trading partner. The questions it raises, particularly with regard to the Executive's power to carry out law enforcement operations abroad, perhaps resonate to a broader audience today than when the case began. In the midst of contemporary anxiety about the struggle against global terrorism, there is a natural concern about the reach and limitations of our political branches in bringing international criminals to justice.

But we need not delve into the legal quagmire of apprehending terrorists or even resolve many of the complex issues spawned by this international abduction dispute. Nor is it within our province to address the policy and diplomatic issues associated with transborder kidnapping. Rather, this appeal presents only the narrow question whether Alvarez has a remedy at law under the ATCA and the FTCA for a violation of the "law of nations."

More precisely, we must determine the statutory authority of a single federal agency — the DEA — to make a warrantless arrest outside the borders of the United States and, if the agency lacks that authority, whether Alvarez has a remedy at law under the ATCA or the FTCA. After a careful review of the relevant statutes, we conclude that the DEA had no authority to effect Alvarez's arrest and detention in Mexico, and that he may seek relief in federal court.

Whatever the contours of the powers of the political branches during wartime or in matters of national security, the exercise of those powers in the combat against terrorism are not implicated in our analysis. Our holding today, that Alvarez may pursue civil remedies for actions taken against him more than ten years ago by the DEA and its agents, is a limited one. It does not speak to the authority of other enforcement agencies or the military, nor to the capacity of the Executive to detain terrorists or other fugitives under circumstances that may implicate our national security interests. The Fourth Circuit recently underscored this distinction when it recognized, in approving the detention of an American citizen captured abroad and designated as an "enemy combatant," that it was "not . . . dealing with a defendant who has been indicted on criminal charges in the exercise of the executive's law enforcement powers" but rather "with the executive's assertion of its power to detain under the war powers of Article II." Hamdi v. Rumsfeld, 316 F.3d 450, 473 (4th Cir. 2003). We, by contrast, are dealing with the former, not the latter.

BACKGROUND

In February 1985, DEA Special Agent Enrique Camarena-Salazar ("Camarena") was abducted and brought to a house in Guadalajara, Mexico, where he was tortured and murdered. Alvarez, a Mexican citizen and a medical doctor who practices in Guadalajara, was present at the house.

Five years after Camarena's death, a federal grand jury in Los Angeles indicted Alvarez for participating in the scheme, and the United States District Court for the Central District of California issued a warrant for his arrest. The United States negotiated with Mexican government officials to take custody of Alvarez, but made no formal request to extradite him. Instead, DEA headquarters in Washington, D.C., approved the use of Mexican nationals, who were not affiliated with either government, to arrest Alvarez in Mexico and to bring him to the United States.

The DEA agent in charge of the Camarena murder investigation, Hector Berellez ("Berellez"), with the approval of his superiors in Los Angeles and Washington, hired Antonio Garate-Bustamante ("Garate"), a Mexican citizen and DEA operative, to contact Mexican nationals who could help apprehend Alvarez. Through a Mexican intermediary, Ignacio Barragan ("Barragan"), Garate arranged for Jose Francisco Sosa ("Sosa"), a former Mexican policeman, to participate in Alvarez's apprehension. Barragan told Sosa that the DEA had obtained a warrant for Alvarez's arrest, would pay the expenses of the arrest operation, and, if the operation was successful, would recommend Sosa for a position with the Mexican Attorney General's Office.

On April 2, 1990, Sosa and others abducted Alvarez from his office and held him overnight at a motel. The next day, they flew him by private plane to El Paso, Texas, where federal agents arrested him. Alvarez was later arraigned and transported to Los Angeles for trial. He remained in federal custody from April 1990 until December 1992.

Alvarez moved to dismiss the indictment, arguing that the federal courts lacked jurisdiction to try him because his arrest violated the United States-Mexico Extradition Treaty. Both the district court and this court agreed, see United States v. Alvarez-Machain ("Alvarez-Machain I"), 946 F.2d 1466, 1466-67 (9th Cir. 1991) (per curiam), aff'g United States v. Caro-Quintero, 745 F.Supp. 599 (C.D.Cal. 1990), but the Supreme Court reversed and remanded the case for trial. See United States v. Alvarez-Machain ("Alvarez-Machain II"), 504 U.S. at 669-70, 112 S.Ct. 2188.

The Supreme Court held that Alvarez's arrest did not violate the United States-Mexico Extradition Treaty. Applying the doctrine announced in Ker v. Illinois, 119 U.S. 436, 7 S.Ct. 225, 30 L.Ed. 421 (1886), the Court held that a court retains its power to try a person for a crime even where the person has been brought within the court's jurisdiction by forcible abduction. Alvarez-Machain II, 504 U.S. at 670, 112 S.Ct. 2188. Significantly, however, the Court noted that Alvarez's abduction "may be in violation of general international law principles" and did not foreclose Alvarez from later pursuing a civil remedy. See id. at 669, 112 S.Ct. 2188; see also Ker, 119 U.S. at 444, 7 S.Ct. 225 (stating that "[t]he [kidnapped] party himself would probably not be without redress, for he could sue [the kidnapper] in an action of trespass and false imprisonment, and the facts set out in the plea would without doubt sustain the action").

Following the Supreme Court's ruling, the case proceeded to trial in 1992. After the presentation of the government's case, the district judge granted a motion for judgment of acquittal on the ground that the government had adduced insufficient evidence to support a guilty verdict. The court concluded that the case against Alvarez was based on "suspicion and . . . hunches but . . . no proof," and that the government's theories were "whole cloth, the wildest speculation."

In 1993, after returning to Mexico, Alvarez filed this action against Sosa, Garate, five unnamed Mexican civilians, the United States, and four DEA agents. The amended complaint alleged a number of conventional and constitutional torts.¹

1 Specifically, Alvarez alleged the following conventional tort claims: (1) kidnapping; (2) torture; (3) cruel, inhuman, and degrading treatment or punishment; (4) arbitrary detention; (5) assault and battery; (6) false imprisonment; (7) intentional infliction of emotional distress; (8) false arrest; (9) negligent employment; and (10) negligent infliction of emotional distress. Alvarez alleged constitutional torts under the Fourth, Fifth, and Eighth Amendments for the acts of kidnapping, torture, cruel and inhuman and degrading treatment or punishment, denial of adequate medical treatment, and arbitrary detention.

The district court substituted the United States for the DEA agents, except Sosa and Garate, on all nonconstitutional claims. The parties later stipulated to the substitution of the United States for Garate. Sosa's interlocutory appeal on the substitution motion was dismissed for lack of appellate jurisdiction. See Alvarez-Machain v. United States ("Alvarez-Machain III"), 107 F.3d 696, 700 n. 2 (9th Cir. 1997) (as amended).

In Alvarez-Machain III, we also affirmed the district court's dismissal of the constitutional claims arising out of harms suffered by Alvarez in Mexico, the denial of the DEA agents' defense based on qualified immunity, and the denial of the United States' defense that the FTCA claims were time-barred. We reversed the district court's dismissal of a claim under the Torture Victims Protection Act of 1991, Pub.L. No. 102-256, 106 Stat. 73. 107 F.3d at 703-04.²

2 The constitutional claims under Bivens v. Six Unknown Named Agents, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971), and the Torture Act claim are no longer at issue.

Upon remand, the district court entered summary judgment for Alvarez on his claims against Sosa for kidnapping and arbitrary detention under the ATCA. The court held that both state-sponsored, transborder abductions and arbitrary detentions violated customary international law.³ The court granted summary judgment to the United States, however, on Alvarez's FTCA claims, concluding that Alvarez's apprehension was privileged and was not a false arrest under California law.

3 The district court found that a third claim brought by Alvarez under the ATCA for cruel, inhuman, and degrading treatment was barred by the law of the case.

These rulings left for resolution the question of Sosa's liability on the remaining tort claims, as well as the calculation of damages on the kidnapping and arbitrary detention claims. After a bench trial, the district court found for Sosa on all remaining claims and held that Alvarez could recover damages under the ATCA only for his detention in Mexico prior to his arrival in the United States. The court applied federal common law, rather than Mexican law, for the calculation of damages and awarded Alvarez $25,000.

These consolidated appeals followed. Sosa appeals the judgment against him, claiming that the district court erred in allowing a cause of action under the ATCA and in applying federal common law, rather than Mexican law, for the calculation of damages. On the ATCA claims, Alvarez appeals the district court's substitution of the United States for the DEA agents and the limitation of damages to those suffered during his imprisonment in Mexico. He also appeals the dismissal of his FTCA claims.

A three-judge panel of this court affirmed Sosa's liability on the ATCA claims, upheld the substitution and damages rulings under ATCA, and reversed the dismissal of Alvarez's FTCA claims. Alvarez-Machain v. United States ("Alvarez-Machain IV"), 266 F.3d 1045, 1064 (9th Cir. 2001), reh'g en banc granted, 284 F.3d 1039, 1040 (9th Cir. 2002).

DISCUSSION

I. ALIEN TORT CLAIMS ACT-JURISDICTION AND CAUSE OF ACTION

The ATCA provides that "[t]he district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States." 28 U.S.C. § 1350. Although enacted in 1789 as part of the first Judiciary Act, the ATCA received little attention until 1980,⁴ when the Second Circuit, in a comprehensive analysis of the statute, held that the ATCA provided subject matter jurisdiction over an action brought by Paraguayan citizens for torture — a violation of the law of nations — committed in Paraguay. See Filartiga v. Pena-Irala (Filartiga I), 630 F.2d 876 (2d Cir. 1980).

4 In 1975, Judge Friendly remarked that the statute had been invoked so rarely since its inception that it existed as "a kind of legal Lohengrin; although it has been with us since the first Judiciary Act . . . no one seems to know whence it came." IIT v. Vencap, Ltd., 519 F.2d 1001, 1015 (2d Cir. 1975) (noting the paucity of cases under the Act and holding that no jurisdiction existed under the Act for fraud and securities claims against foreign corporations).

Since the Filartiga I decision, the ATCA has been invoked in a variety of actions alleging human rights violations. See, e.g., Abebe-Jira v. Negewo, 72 F.3d 844 (11th Cir. 1996) (affirming judgment under ATCA against former Ethiopian official for torture and cruel, inhuman, and degrading treatment); Kadic v. Karadzic, 70 F.3d 232 (2d Cir. 1995) (concluding that alleged war crimes, genocide, torture, and other atrocities committed by a Bosnian Serb leader were actionable under the ATCA); Tel-Oren v. Libyan Arab Republic, 726 F.2d 774 (D.C. Cir. 1984) (dismissing for lack of subject matter jurisdiction claims brought against the Palestine Liberation Organization, the Libyan government, and other entities for terrorist activities allegedly in violation of the law of nations); Xuncax v. Gramajo, 886 F.Supp. 162 (D.Mass. 1995) (deeming torture, summary execution, "disappearance," and arbitrary detention by Guatemalan military to be actionable violations under the ATCA).

Our first opportunity to address the scope of the ATCA came in Trajano v. Marcos (In re Estate of Marcos Human Rights Litig.) ("Marcos I"), 978 F.2d 493 (9th Cir. 1992), a wrongful death action against former Philippine President Ferdinand Marcos and his daughter for the torture and murder of a Philippine citizen. We recognized that "it would be unthinkable to conclude other than that acts of official torture violate customary international law," and concluded that the plaintiff, an alien, had properly invoked the subject matter jurisdiction of the federal courts under the ATCA. Id. at 499 (citation and internal quotation marks omitted). Referencing an April 1787 letter from James Madison to Edmond Randolph, we concluded that "[t]here is ample indication that the 'Arising Under' Clause was meant to extend the judicial power of the federal courts . . . to 'all cases which concern foreigners.'" Id. at 502. Because the "Arising Under" Clause gave Congress the power to enact the ATCA, we held that exercising jurisdiction over the claims would not run afoul of Article III of the Constitution. Id. at 502-03.

When the Marcos litigation returned to this court in Hilao v. Estate of Marcos (In re Estate of Marcos, Human Rights Litig.) ("Marcos II"), 25 F.3d 1467 (9th Cir. 1994), we further delineated the contours of the ATCA.⁵ We resolved that the Act not only provides federal courts with subject matter jurisdiction, but also creates a cause of action for an alleged violation of the law of nations: "[S]ection 1350 does not require that the action 'arise under' the law of nations, but only mandates a 'violation of the law of nations' in order to create a cause of action." Id. at 1475 (quoting Tel-Oren, 726 F.2d at 779 (Edwards, J., concurring)). In other words, "[n]othing more than a violation of the law of nations is required to invoke section 1350." Id. (citation omitted).

5 Following Marcos II, we issued several other decisions in relation to the Marcos litigation, two of which are referenced in this opinion: Hilao v. Estate of Marcos ("Marcos III"), 103 F.3d 767 (9th Cir. 1996) and Hilao v. Estate of Marcos ("Marcos IV"), 103 F.3d 789 (9th Cir. 1996).

Of course, not every violation of international law constitutes an actionable claim under the ATCA. In Marcos II, we were careful to limit actionable violations to those international norms that are "specific, universal, and obligatory." Id. at 1475. This formulation, which lays the foundation for our approach to international norms, is in keeping with the narrow scope of ATCA jurisdiction and the general practice of limiting judicial review to those areas of international law that have achieved sufficient consensus to merit application by a domestic tribunal. See Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 428, 84 S.Ct. 923, 11 L.Ed.2d 804 (1964) ("[T]he greater the degree of codification or consensus concerning a particular area of international law, the more appropriate it is for the judiciary to render decisions regarding it. . . ."); cf. United States v. Smith, 18 U.S. (5 Wheat.) 153, 162, 5 L.Ed. 57 (1820) (finding piracy "universally treat[ed] . . . as an offence against the law of nations" and "sufficiently and constitutionally defined" by commentators to be punishable by Congress).

Sosa urges a narrow reading of the "law of nations" and a correspondingly strict interpretation of the "specific, universal, and obligatory" requirement. He argues that only violations of jus cogens norms, as distinguished from violations of customary international law, are sufficiently "universal" and "obligatory" to be actionable as violations of "the law of nations" under the ATCA. We decline to embrace this restrictive reading, as we are guided by the language of the statute, not an imported restriction.

The term jus cogens refers to a category of "peremptory norms" that are "'accepted and recognized by the international community of states as a whole as . . . norm[s] from which no derogation is permitted.'" Siderman de Blake v. Republic of Argentina, 965 F.2d 699, 714 (9th Cir. 1992) (quoting Vienna Convention on the Law of Treaties, art. 53, May 23, 1969, 1155 U.N.T.S. 332, 8 I.L.M. 679). Customary international law, a direct descendent of the "law of nations," is a related, but distinct, concept. Id. It refers more generally to those established norms of contemporary international law that are "ascertain[ed] . . . 'by consulting the works of jurists, writing professedly on public law; or by the general usage and practice of nations; or by judicial decisions recognizing and enforcing that law.'" Id. at 714-15 (quoting Smith, 18 U.S. at 160-61).

We have explained the difference between these two concepts as follows:

While jus cogens and customary international law are related, they differ in one important respect. Customary international law, like international law defined by treaties and other international agreements, rests on the consent of states. A state that persistently objects to a norm of customary international law that other states accept is not bound by that norm

. . . .

In contrast, jus cogens embraces customary laws considered binding on all nations and is derived from values taken to be fundamental by the international community, rather than from the fortuitous or selfinterested choices of nations. Whereas customary international law derives solely from the consent of states, the fundamental and universal norms constituting jus cogens transcend such consent. . . .

Because jus cogens norms do not depend solely on the consent of states for their binding force, they enjoy the highest status within international law.

Id. at 715 (internal quotation marks and citations omitted).⁶

6 The commentators embrace this distinction. See 1 M. Cherif Bassiouni, International Criminal Law 40 (2d ed. 1999) ("[A] jus cogens norm holds the highest hierarchical position among all other norms and principles. As a consequence of that standing, jus cogens norms are deemed to be 'peremptory' and 'non-derogable.'"); Ian Brownlie, Principles of Public International Law 515 (5th ed. 1998) ("The major distinguishing feature of [ jus cogens] rules is their relative indelibility.").

Given the non-derogable nature of jus cogens norms, it comes as no surprise that we have found that a jus cogens violation is sufficient to satisfy the "specific, universal, and obligatory" standard. See Marcos II, 25 F.3d at 1475. But the fact that a violation of this subcategory of international norms is sufficient to warrant an actionable claim under the ATCA does not render it necessary. Indeed, our recent cases lay out the components of an actionable violation without reference to jus cogens. See Papa v. United States, 281 F.3d 1004, 1013 (9th Cir. 2002) (remanding case to district court to apply the "applicable standard," which requires plaintiffs to allege "specific, universal, and obligatory" norms as part of their claim); Martinez v. City of Los Angeles, 141 F.3d 1373, 1383-84 (9th Cir. 1998) (recognizing, without a discussion of jus cogens, that arbitrary detention meets the standard for a cognizable ATCA claim).

The notion of jus cogens norms was not part of the legal landscape when Congress enacted the ATCA in 1789. See Brownlie, supra, at 516 (explaining the modern evolution of jus cogens). Thus, to restrict actionable violations of international law to only those claims that fall within the categorical universe known as jus cogens would deviate from both the history and text of the ATCA.

Although a strict categorical approach may have surface appeal for its apparent ease of application, it is far from certain which norms would qualify for jus cogens status. The development of an elite category of human rights norms is of relatively recent origin in international law, and "[a]lthough the concept of jus cogens is now accepted, its content is not agreed." Restatement (Third) of the Foreign Relations Law of the United States § 102 n. 6 (1987) ("Restatement on Foreign Relations"). As one respected commentator put it, "more authority exists for the category of jus cogens than exists for its particular content. . . ." Brownlie, supra, at 516-17; see also Theodor Meron, On a Hierarchy of International Human Rights, 80 A.J.I.L. 1, 14-15 (1986) (explaining the difficulties of strict categorization in defining peremptory norms). We therefore remain confident that the standard established in Marcos II and repeated throughout our case law best reflects the text and purpose of the ATCA and provides sufficient guidance for evaluating Alvarez's claim.

With this international law background in mind, we turn to Alvarez's contentions on appeal. Alvarez argues that he has a remedy under the ATCA for two separate violations of international law. First, he claims that state-sponsored abduction within the territory of another state without its consent is a violation of the international law of sovereignty and the customary norms of international human rights law. Second, he contends that his seizure and confinement violated the international customary legal norm against arbitrary arrest and detention.

In view of the dissent's rhetoric and lengthy discourse, it may not be readily apparent that the dissent is in accord with a significant portion of our holding. Ten members of the en banc court agree that Alvarez lacks standing to obtain redress for claims based on an alleged violation of Mexico's sovereignty and that his claim for transborder abduction fails.⁷ These same judges also agree that there is a universally recognized norm prohibiting arbitrary arrest and detention. It is only as to the application of this latter norm that we part company.

7 Judge Gould's solitary dissent on the political question issue misses the mark, as the other dissenters acknowledge. See infra at n. 2 (O'Scannlain, J., dissenting). The mere fact that this case raises difficult and politically sensitive issues connected to our foreign relations does not preclude us from carrying out the legislative mandate of Congress under § 1350. See Baker v. Carr, 369 U.S. 186, 211, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962) ("[I]t is error to suppose that every case or controversy which touches upon foreign relations lies beyond judicial cognizance."). The crux of the claim here rests on legislative delegation, not foreign relations. We see a critical distinction between, on the one hand, second guessing the foreign policy judgments of the political branches to whom such judgments have been constitutionally assigned and, on the other hand, reviewing claims based in tort and brought under federal statutes instructing the judiciary to adjudicate such claims. See Kadic, 70 F.3d at 249 ("The department to whom this [tort suit against the PLO] has been constitutionally committed is none other than our own — the Judiciary.") (internal quotation marks and citation omitted); Abebe-Jira, 72 F.3d at 848 (holding that the political question doctrine did not bar tort action brought by former prisoners in Ethiopia under the ATCA).

A. TRANSBORDER ABDUCTION AND THE LAW OF NATIONS

1. STANDING AND SOVEREIGNTY

Alvarez claims that his arrest violated Mexico's sovereign rights because Mexico had not granted the United States permission to exercise police power on its soil. Because such an encroachment on Mexico's sovereignty violates "the law of nations" within the meaning of the ATCA, Alvarez reasons, he is entitled to relief under that statute. The district court agreed and rejected Sosa's objection that Alvarez lacks standing to invoke Mexico's sovereignty rights.

We have little trouble accepting the premise from which Alvarez begins. Few principles in international law are as deeply rooted as the general norm prohibiting acts of sovereignty that offend the territorial integrity of another state. See generally 1 L. Oppenheim, Oppenheim's International Law § 119 (Robert Jennings Arthur Watts eds., 9th ed. 1992); see also F.A. Mann, Reflections on the Prosecution of Persons Abducted in Breach of International Law, in International Law at a Time of Perplexity 407 n. 2 (Yoram Dinstein Mala Tabory eds. 1989) (referring to this "incontrovertible" rule as "elementary"). This tenet, as Alvarez points out, can be traced to the earliest decisions of the Supreme Court. Most notably, in 1812, when faced with the question whether an American citizen could assert title to an armed French vessel found in the territorial waters of the United States, Justice Marshall began his landmark decision by emphasizing the "exclusive and absolute" nature of territorial jurisdiction, exceptions to which "must be traced up to the consent of the nation itself." Schooner Exchange v. M'Faddon, 11 U.S. (7 Cranch) 116, 136, 3 L.Ed. 287 (1812).

Twelve years later, Justice Story voiced similar sentiments. Analyzing an American seizure of a foreign ship that had sailed into Spanish waters, he observed that "[i]t would be monstrous to suppose that our . . . officers were authorized to enter into foreign ports and territories, for the purpose of seizing vessels which had offended against our laws. It cannot be presumed that Congress would voluntarily justify such a clear violation of the laws of nations." The Apollon, 22 U.S. (9 Wheat.) 362, 371, 6 L.Ed. 111 (1824).

Alvarez seeks to invoke a principle, concomitant with this precept of territorial sovereignty, that prohibits a state's law enforcement agents from exercising their functions in the territory of another state without the latter's consent. The Supreme Court clearly recognized this proscription in The Appollon. In addition, several notable authorities are in accord. See Restatement on Foreign Relations § 432(2) ("A state's law enforcement officers may exercise their functions in the territory of another state only with the consent of the other state, given by duly authorized officials of that state."); 1 Oppenheim, supra, § 119, at 387-88 ("It is . . . a breach of international law for a state without permission to send its agents into the territory of another state to apprehend persons accused of having committed a crime."); see also M. Cherif Bassiouni, International Extradition: United States Law and Practice 255 (4th ed. 2002) (recognizing the rule and noting that it is "grounded in the notion that international law is designed to protect the sovereignty and territorial integrity of states by restricting impermissible state conduct"). But whatever the modern contours of this principle or its corollaries, they are inapplicable here and need not be explored because Alvarez cannot establish, as a threshold matter, that he has standing to assert Mexico's interests in its territorial sovereignty.⁸

8 Although we need not examine the place of such a rule in customary international law or as it applies to this case, we note that Alvarez's assertion is not wholly straightforward, as it raises complex questions about the intersection of extraterritorial criminal jurisdiction, extraterritorial enforcement, and state sovereignty. The three concepts are not necessarily correlative as a matter of international law. See, e.g., S.S. Lotus (Turk. v. Fr.), 1927 P.C.I.J. (ser. A) No. 10, at 19 (Sept. 7) ("The territoriality of criminal law . . . is not an absolute principle of international law and by no means coincides with territorial sovereignty."). And although extraterritoriality is well-established in our jurisprudence, see infra Part I.B., to the extent that either extraterritorial jurisdiction or extraterritorial enforcement overlap with the national laws and policies of another state, inevitably there is a potential for friction between states. See Bassiouni, International Extradition, supra, at 314 n. 1.

The Supreme Court has instructed that to meet the "irreducible constitutional minimum of standing" under Article III, plaintiffs must "[f]irst and foremost" show the existence of an "injury in fact." Steel Co. v. Citizens for a Better Env't, 523 U.S. 83, 102-03, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998) (internal citations and quotation marks omitted). Related to this constitutional prerequisite is a separate "prudential" requirement of standing: plaintiffs must demonstrate they are "proper proponents of the particular legal rights on which they base their suit." Singleton v. Wulff, 428 U.S. 106, 112, 96 S.Ct. 2868, 49 L.Ed.2d 826 (1976). This requirement applies "even when the very same allegedly illegal act that affects the litigant also affects a third party." United States Dep't of Labor v. Triplett, 494 U.S. 715, 720, 110 S.Ct. 1428, 108 L.Ed.2d 701 (1990). Although Alvarez may have properly alleged that Mexico's sovereignty was infringed during his abduction — an issue we need not resolve here — he has not demonstrated that he is a proper party to vindicate Mexico's national interests.

Alvarez argues that he meets the standing requirements because courts may review ATCA claims whenever an alien "is injured tortiously in the course of the defendant's violation of international law." But the ATCA creates a remedy for "a tort . . . committed in violation of the law of nations," not "in the course of" any recognized international law violation. 28 U.S.C. § 1350. The legal rights on which Alvarez bases his claim, and which the ATCA recognizes, are those that protect the individual from tortious conduct. By its terms, the ATCA provides only for suits by individual aliens; it does not allow for an individual to vindicate the rights of a foreign government.

To allow state-on-state injuries like the one Alvarez alleges here to be vindicated by a third party not only would read too much into the ATCA, but would lead to the judiciary's intrusion into matters that are appropriately reserved for the Executive branch. Although international human rights litigation under the ATCA inevitably raises issues implicating foreign relations, sovereigns' prerogatives are ordinarily and traditionally handled through diplomatic channels.⁹ The right of a nation to invoke its territorial integrity does not translate into the right of an individual to invoke such interests in the name of the law of nations.

9 We do not mean to imply that an individual never has a claim for breach of the law of nations for which a state-to-state remedy also exists. See Restatement on Foreign Relations § 703(1) (establishing states' rights to take action against fellow states that transgress international human rights norms). We note, however, that the commentary of the Restatement on Foreign Relations indicates that most state-to-state remedies are subordinated to individual remedies where transgressor states' domestic law makes such remedies available. See id. §§ 703 cmt. d, 713 cmt. f.

Alvarez seeks refuge in Ker v. Illinois, 119 U.S. 436, 7 S.Ct. 225, 30 L.Ed. 421 (1886), the case that previously doomed his attempt to secure dismissal of his criminal indictment. See Alvarez-Machain II, 504 U.S. at 662, 112 S.Ct. 2188. Like Alvarez, Ker claimed forcible abduction from a foreign country, in his case Peru. Although the Supreme Court refused to dismiss Ker's indictment, it observed that Ker was "probably not . . . without redress, for he could sue [his abductor] in an action of trespass and false imprisonment." Ker, 119 U.S. at 444, 7 S.Ct. 225. The Court made no guarantees, however, regarding a claim under the ATCA or any other federal statute; nor did it intimate that Ker could sue to avenge Peru's sovereignty rights. Rather, the Court noted that Peru could pursue a separate remedy — the kidnapper's extradition. Id. Ker thus implicitly drew the distinction between vindication of individual rights and a sovereign's vindication of its rights. Ker does not bridge the gap in Alvarez's claim.

2. TRANSBORDER ABDUCTION AND CUSTOMARY INTERNATIONAL LAW

Apparently cognizant of the constitutional barrier to his claim, Alvarez offers an alternative theory: he seeks to bypass the standing hurdle by arguing that, notwithstanding any infringements upon Mexico's sovereignty, the act of transborder kidnapping was, in itself, a violation of customary international human rights law. This norm, as defined by Alvarez, creates a personal right under the law of nations.

Sosa, the DEA agents, and the United States all urge that this norm fails the law of nations test. They contend that the prohibition that Alvarez identifies has not reached the level of acceptance in the international community sufficient to qualify as "universal" and "obligatory." They also argue that, whatever degree of agreement other nations have reached, the United States has affirmatively and definitively rejected this principle. We agree. The United States does not recognize a prohibition against transborder kidnapping, nor can it be said that there is international acceptance of such a norm.

We embrace the Supreme Court's directive that the law of nations "may be ascertained by consulting the work of jurists, writing professedly on public law; or by the general usage and practice of nations; or by judicial decisions recognising and enforcing that law." Smith, 18 U.S. at 160-61; see also The Paquete Habana, 175 U.S. 677, 700, 20 S.Ct. 290, 44 L.Ed. 320 (1900) ("[W]here there is no treaty, and no controlling executive or legislative act or judicial decision, resort must be had to the customs and usages of civilized nations; and, as evidence of these, to the works of jurists and commentators. . . ."). Evidence of the law of nations may also be garnered from international agreements and United Nations declarations. See Siderman, 965 F.2d at 716-17; Filartiga I, 630 F.2d at 883-84.

Article 38 of the Statute of the International Court of Justice serves as a convenient summary of the sources of international law, although we recognize that defining "[t]he 'sources' of international law is a subject of much continuing scholarship." United States v. Yousef, 327 F.3d 56, 100-01 (2d Cir. 2003). Article 38 provides, in part:

1. The Court, whose function is to decide in accordance with international law such disputes as are submitted to it, shall apply:

a. international conventions, whether general or particular, establishing rules expressly recognized by the contesting states;

b. international custom, as evidence of a general ractice accepted as law;

c. the general principles of law recognized by civilized nations;

d. subject to the provisions of Article 59, judicial decisions and the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law.¹⁰

10 Article 59 states: "The decision of the Court has no binding force except between the parties and in respect of that particular case."

Statute of the International Court of Justice, June 26, 1945, art. 38, 59 Stat. 1055, U.S.T.S. 993.

International agreements to which the United States is a signatory provide an obvious and convenient starting point. It would be, of course, a relatively simple analysis if we could pinpoint in such an agreement a prohibition against transborder abductions. Despite eloquent arguments to the contrary, we find no such support in the text of any international agreement.

Alvarez and the amici point to a number of international human rights instruments which, they argue, support an individual right to remain free of transborder abductions. But no authority cited by Alvarez recognizes an explicit prohibition against forcible abduction.¹¹ Rather, each of the authorities speaks to general prohibitions against restricting an individual's right to freedom and movement and security of person. For example, the American Convention on Human Rights ("American Convention"), which Alvarez cites, states that "[e]very person has the right to personal liberty and security" and "[n]o one shall be deprived of his physical liberty except for the reasons and under the conditions established beforehand by the constitution of the State Party concerned or by a law established pursuant thereto." Art. 7(1), 7(2), opened for signature Nov. 22, 1969, 1144 U.N.T.S. 123 (signed but not ratified by the United States). Similarly, the International Covenant on Civil and Political Rights ("ICCPR") provides that "[e]veryone lawfully within the territory of a State shall, within that territory, have the right to liberty of movement and freedom to choose his residence." Art. 12, G.A. Res. 2200, 21 U.N. GAOR, Supp. No. 16, at 52, U.N. Doc. A/6316 (1966), 999 U.N.T.S. 171 (entered into force Mar. 23, 1976) (ratified by the United States Sept. 8, 1992). See also Universal Declaration of Human Rights ("Universal Declaration"), art. 13(1), G.A. Res. 217A (III), 3 U.N. GAOR, Supp. No. 16, U.N. Doc. A/810 (1948) ("Everyone has the right to freedom of movement and residence within the borders of each state.");¹² American Declaration of the Rights and Duties of Man, art. VIII, May 2, 1948, O.A.S. Res. XXX, reprinted in Basic Documents Pertaining to Human Rights in the Inter-American System, OEA/Ser.LV/II. 82 doc. 6 rev. 1, at 17 (1992) ("Every person has the right to fix his residence within the territory of the state of which he is a national, to move about freely within such territory, and not to leave it except by his own will."). Such general prohibitions are insufficient to support Alvarez's claim that there is an international norm against transborder abduction because an actionable claim under the ATCA requires the showing of a violation of the law of nations that is "specific, universal, and obligatory."

11 The Restatement on Foreign Relations reflects this void: "None of the international human rights conventions to date . . . provides that forcible abduction or irregular extradition is a violation of international human rights law." Restatement on Foreign Relations § 432 n. 1.

12 We have recognized that the Universal Declaration, although not binding on states, constitutes "a powerful and authoritative statement of the customary international law of human rights." Siderman, 965 F.2d at 719.

Looking beyond the declarations and covenants to treaties does not yield a different result.¹³ At the time of Alvarez's abduction, the United States-Mexico Extradition Treaty did not extend to transborder abduction and there was no separate treaty with such a prohibition. See Alvarez-Machain II, 504 U.S. at 669-70, 112 S.Ct. 2188. The absence of any agreement is consistent with our conclusion that the United States has not embraced the prohibition urged by Alvarez. That is not to say that Alvarez's abduction went unnoticed. Indeed, it was met with a formal diplomatic protest by Mexico and considerable public outcry.¹⁴

13 The ATCA permits suits for both a "violation of the law of nations" and torts in violation of "a treaty of the United States." 28 U.S.C. § 1350.

14 The Mexican Government filed an official protest with the United States, presenting a diplomatic note to the U.S. Department of State on three separate occasions. See Brief for the United Mexican States as Amicus Curiae in Support of Affirmance at 3-4, Alvarez-Machain II, reprinted in 31 I.L.M. 934, 938-39 (1992); see also Caro-Quintero, 745 F.Supp. at 604. The resulting friction between the United States and Mexico was well documented. See, e.g., Marjorie Miller Douglas Jehl, Mexico to Confront U.S. on Camarena Case Abduction, L.A. Times, April 18, 1990, at A1; Carlyle C. Douglas, Arm of U.S. Law Is Too Long, Mexico Complains, N.Y. Times, April 22, 1990, § 4, at 11; Jack Epstein, Growing Uproar in Mexico About Alleged Abuses by U.S., S.F. Chron., July 7, 1992, at A8.

In 1994, four years after Alvarez was abducted, the United States and Mexico reached an agreement to prohibit the practice of transborder arrest. Treaty to Prohibit Transborder Abductions, Nov. 23, 1994, U.S.-Mex., reprinted in Michael Abbell, Extradition to and From the United States, at A-303 (2002). That agreement is not yet in force, however, because the President has not submitted it to the Senate for its advice and consent. See id. at A-287. In any event, the proposed treaty would not help Alvarez: it would explicitly foreclose the right of abductees to sue their abductors. See id. at A-303. If anything, this development underscores the void that existed before the treaty was signed and the reality that the United States does not yet consider itself bound by the supposed norm against transborder abductions. Alvarez offers no other legislative or judicial source that supports a specific, enforceable norm against transborder abductions.

The United States claims that unilateral, transborder abductions are a "rare" occurrence. And the notion of sneaking across the border to nab a criminal suspect surely raises more than a diplomatic eyebrow. Nonetheless, our review of the international authorities and literature reveals no specific binding obligation, express or implied, on the part of the United States or its agents to refrain from transborder kidnapping. Nor can we say that there is a "universal" consensus in the sense that we use that term to describe well-entrenched customs of international law. Any agreement that may exist on this score has failed to surface in the declarations and accords that commonly manifest the mutual concern of states. See Filartiga I, 630 F.2d at 888 ("It is only where the nations of the world have demonstrated that the wrong is of mutual, and not merely several, concern, by means of express international accords, that a wrong generally recognized becomes an international law violation within the meaning of the statute."). Because a human rights norm recognizing an individual's right to be free from transborder abductions has not reached a status of international accord sufficient to render it "obligatory" or "universal," it cannot qualify as an actionable norm under the ATCA. This is a case where aspiration has not yet ripened into obligation.¹⁵ B. ARBITRARY ARREST AND DETENTION AND THE LAW OF NATIONS

15 The dissent asserts that we could shortcut our analysis and make ATCA review "easier" by determining, as a threshold matter, whether the United States, through the political branches, has decided variously not to "recognize," "assent," "agree with," or "subscribe to" an international norm prohibiting transborder arrests. Should the United States demonstrate any form of non-acquiescence, the customary international law norm would, according to the dissent, fail to achieve "universal" status for purposes of ATCA liability.
Although we accept the well-established principle that customary norms are fundamentally based on the consent of states, and that the United States might well decide to deliberately disavow or repudiate certain principles of international law, we cannot agree with the dissent's implication that every executive branch decision to breach an international norm translates into a more global repudiation of that norm or necessarily insulates the United States and its agents from civil tort liability. Our understanding accords with Ker, 119 U.S. at 444-45, 7 S.Ct. 225 (holding that civil remedies might still be available for violations of treaties or the law of nations even though jurisdiction to prosecute a defendant criminally may not be invalidated by an extraterritorial abduction), and the Supreme Court's more recent acknowledgment that Alvarez might be correct that his abduction was "shocking" and "in violation of general international law principles" Alvarez-Machain II, 504 U.S. at 669, 112 S.Ct. 2188; see also The Paquete Habana, 175 U.S. 677, 700, 20 S.Ct. 290, 44 L.Ed. 320 (1900) (stating that "[i]nternational law is part of our law," and that "where there is no treaty, and no controlling executive or legislative act or judicial decision, resort must be had to the customs and usages of civilized nations; and, as evidence of these, to the works of jurists and commentators. . . .") (emphasis added); Yousef, 327 F.3d at 92 n. 25 ("While it is not possible to claim that the practice or policies of any one country, including the United States, has any such authority that the contours of customary international law may be determined by reference only to that country, it is highly unlikely that a purported principle of customary international law in direct conflict with the recognized practices and customs of the United States and/or other prominent players in the community of States could be deemed to qualify as a bona fide customary international law principle.") (emphasis added); Louis Henkin, Foreign Affairs and the U.S. Constitution 243 (2d ed. 1996) (explaining that "[u]nlike treaties . . . principles of customary international law cannot be denounced or terminated by the President and cannot be eliminated from the law of the United States by any Presidential act.") (emphasis added).

Alvarez is not, however, without a remedy. The unilateral, nonconsensual extraterritorial arrest and detention of Alvarez were arbitrary and in violation of the law of nations under the ATCA.

1. THE PROHIBITION AGAINST ARBITRARY ARREST AND DETENTION

Unlike transborder arrests, there exists a clear and universally recognized norm prohibiting arbitrary arrest and detention. This prohibition is codified in every major comprehensive human rights instrument and is reflected in at least 119 national constitutions. See M. Cherif Bassiouni, Human Rights in the Context of Criminal Justice: Identifying International Procedural Protections and Equivalent Protections in National Constitutions, 3 Duke J. Comp. Int'l L. 235, 260-61 (1993). The Universal Declaration, perhaps the most well-recognized explication of international human rights norms, provides that "[n]o one shall be subjected to arbitrary arrest, detention, or exile," Universal Declaration, art. 9, and the ICCPR, which the United States has ratified,¹⁶ unequivocally obliges states parties to refrain from "arbitrary arrest or detention." ICCPR, art. 9.¹⁷

16 The ICCPR is one of several international covenants designed to formally codify many of the rights embodied in the Universal Declaration. See Brownlie, supra, at 576.

17 Each of the regional human rights instruments contains a similar prohibition. See American Convention, art. 7(3) ("No one shall be subject to arbitrary arrest or imprisonment."); European Convention for the Protection of Human Rights and Fundamental Freedoms ("European Convention"), art. 5(1), opened for signature Nov. 4, 1950, 213 U.N.T.S. 222 (deprivation of liberty must be "in accordance with a procedure prescribed by law" and only in the case of, inter alia, "the lawful arrest or detention of a person effected for the purpose of bringing him before the competent legal authority. . . ."); African Charter on Human and Peoples' Rights ("African Charter"), art. 6, June 27, 1981, 21 I.L.M. 58 (1982) ("[N]o one may be arbitrarily arrested or detained.").

We recently reaffirmed the universal, obligatory, and specific nature of this norm in Martinez, 141 F.3d at 1384 (recognizing a "clear international prohibition against arbitrary arrest and detention"); see also Marcos IV, 103 F.3d at 795 (recognizing "arbitrary detention . . . as[an] actionable violation of international law"). We explained, in defining the norm, that "[d]etention is arbitrary 'if it is not pursuant to law; it may be arbitrary also if it is incompatible with the principles of justice or with the dignity of the human person.'" Martinez, 141 F.3d at 1384 (quoting Restatement on Foreign Relations § 702 cmt. h).¹⁸

18 Our standard reflects the language of the Restatement as well as other major international sources. See Restatement on Foreign Relations § 702 cmt. h; ICCPR, art. 9(1) ("No one shall be subjected to arbitrary arrest or detention. No one shall be deprived of his liberty except on such grounds and in accordance with such procedures as are established by law."); id., art. 9(5) ("Anyone who has been the victim of unlawful arrest or detention shall have an enforceable right to compensation."); European Convention, art. 5(1) (deprivation of liberty must be "in accordance with a procedure prescribed by law" and only in the case of, inter alia, "the lawful arrest or detention of a person effected for the purpose of bringing him before the competent legal authority. . . ."); African Charter, art. 6 ("No one may be deprived of his freedom except for reasons and conditions previously laid down by law. In particular, no one may be arbitrarily arrested or detained."); see also Winterwerp v. Netherlands, 33 Eur. Ct. H.R. (ser. A.) at para. 39 (1979) ("[N]o detention that is arbitrary can ever be regarded as lawful."); United Nations, Study of the Right of Everyone to be Free from Arbitrary Arrest, Detention, and Exile 7 (1964) ("United Nations Study") (adopting the view that "an arrest or detention is arbitrary if it is (a) on grounds or in accordance with procedures other than those established by law, or (b) under the provisions of a law the purpose of which is incompatible with the respect for the right to liberty and security of person").

Sosa acknowledges the prohibition against arbitrary arrest and detention, but he contends that for ATCA liability to attach, Alvarez's detention must be "prolonged" in addition to being arbitrary. We can divine no such requirement in our precedent or in the applicable international authorities. Rather, as the language of the international instruments demonstrates, the norm is universally cited as one against "arbitrary" detention and does not include a temporal element. Other authorities reflect this understanding. See, e.g., Bassiouni, Human Rights in the Context of Criminal Justice, supra, at 260; Paul Sieghart, The International Law of Human Rights 135-59 (1983); see also United Nations Study, supra, at 5-8 (defining elements of the norm without mention of a temporal component).¹⁹

19 This reading is also supported in the case law. See, e.g., de Sanchez v. Banco Central de Nicaragua, 770 F.2d 1385, 1397 (5th Cir. 1985) (recognizing "the right not to be arbitrarily detained" as part of the law of nations); Rodriguez-Fernandez v. Wilkinson, 654 F.2d 1382, 1388 (10th Cir. 1981) ("No principle of international law is more fundamental than the concept that human beings should be free from arbitrary imprisonment."); Paul v. Avril, 901 F.Supp. 330, 333-34, 335 (S.D.Fla. 1994) (concluding plaintiff suffered arbitrary detention although he was held for less than ten hours); Forti v. Suarez-Mason, 672 F.Supp. 1531, 1541 (N.D.Cal. 1987) ("There is case law finding sufficient consensus to evince a customary international human rights norm against arbitrary detention. The consensus is even clearer in the case of a state's prolonged arbitrary detention of its own citizens." (internal citations omitted)); see also Litwa v. Poland, App. No. 26629/95, 33 Eur. H.R. Rep. 53 (2000) (finding detention of six hours and thirty minutes constitutes violation under Article 5 of the European Convention); Quinn v. France, App. No. 18580/91, 21 Eur. H.R. Rep. 529 (1995) (finding claim of arbitrary detention under Article 5 of the European Convention where petitioner was detained for a period of eleven hours).

Although § 702 of the Restatement on Foreign Relations includes a reference to "prolonged arbitrary detention,"²⁰ neither the Restatement nor our cases import a separate temporal requirement for purposes of ATCA liability. Section 702 contains a short list of human rights norms that it deems sufficient to qualify as customary law violations. See Restatement on Foreign Relations § 702(a)-(g). But the comments to § 702 clarify that the list is non-exhaustive and that virtually all of the norms listed, including "prolonged arbitrary detention," belong among the elite set of jus cogens norms that are non-derogable. Id. cmts. a, n. Section 702 does not state that every arbitrary detention must be "prolonged" to qualify as a violation of the law of nations — which is all that is required under the ATCA — and in fact implies the opposite. See id. cmt. ("A single, brief, arbitrary detention by an official of a state party to one of the principal international agreements might violate that agreement."). Likewise, our holding in Martinez, which cited the Restatement, included the length of detention as but one factor among many in determining whether a violation of the law of nations had occurred. 141 F.3d at 1384.

20 The Restatement provides that "[a] state violates international law if . . . it practices, encourages, or condones . . . prolonged arbitrary detention." Restatement on Foreign Relations § 702(e).

This is not to say that the length of detention cannot be a factor in evaluating whether there was an actionable violation of international law. Indeed, an extended detention following an improper arrest would necessarily contribute to "arbitrariness." We simply hold, consistent with international law, that there is no free-standing temporal requirement nor any magical time period that triggers the norm.

2. APPLICATION OF ARBITRARY ARREST AND DETENTION STANDARD TO ALVAREZ

The standard then is whether the arrest and detention were arbitrary, that is, "not pursuant to law."²¹ Martinez,

21 Although the norm against arbitrary arrest and detention may encompass both illegal and unjust acts, we need not decide here under what circumstances an "unjust" arrest or detention might qualify as "arbitrary." See, e.g., Restatement of Foreign Relations § 702 n. 6 ("Detention is arbitrary if it is unlawful or unjust."); Laurent Marcoux, Jr., Protection from Arbitrary Arrest and Detention Under International Law, 5 B.C. Int'l Comp. L. Rev. 345 (1982) (analyzing the language and drafting history of the Universal Declaration and ICCPR as evidence that the term "arbitrary" was chosen to encompass a broader standard than mere unlawfulness).

141 F.3d at 1384. In the case before us, there was, quite simply, no basis in law for the unilateral extraterritorial arrest and related detention of Alvarez in Mexico.

The only instrument Sosa can point to as evidence that Alvarez's abduction was "pursuant to law" is an arrest warrant issued by the United States District Court for the Central District of California. But a federal arrest warrant, without more, hardly serves as a license to effectuate arrests worldwide. It is no accident that the warrant is directed to "The United States Marshal and any Authorized United States Officer" (emphasis added). The Federal Rules of Criminal Procedure in effect at the time of Alvarez's arrest provided that "[a] warrant may be executed . . . within the jurisdiction of the United States." Fed.R.Crim.P. 4(d)(2).²² The language could hardly be clearer — "within the jurisdiction of the United States" means exactly what it says.²³

22 Rule 4(d)(2) was amended on December 1, 2002. The Rule, renumbered as 4(c)(2), now reads, "A warrant may be executed, or a summons served, within the jurisdiction of the United States or anywhere else a federal statute authorizes an arrest" (underscoring indicates amendment). The advisory committee notes clarify that the "new language . . . reflects the recent enactment of the Military Extraterritorial Jurisdiction Act ( Pub.L. No. 106-523, 114 Stat. 2488) that permits arrests of certain military and Department of Defense personnel overseas. See also 14 U.S.C. § 89 (Coast Guard authority to effect arrests outside territorial limits of United States)." Fed.R.Crim.P. 4 advisory committee's note. The calibration of Rule 4 to statutes in which Congress has made explicit the territorial reach of the arrest power demonstrates not only the limited scope of a traditional arrest warrant, but Congress's own recognition that it must speak clearly when expanding the geographical scope of an agent's extraterritorial arrest authority.

23 Alvarez, of course, was only one of many charged in connection with Camarena's murder. An indictment issued on January 30, 1985 charged twenty-two persons with crimes in connection with Camarena's murder. Seven were tried in federal court. Including Alvarez, three of the seven were brought "by means of covert forcible abduction from their homelands." Caro-Quintero, 745 F.Supp. at 602. Alvarez's abduction was unique in that it involved neither the cooperation of local police nor the consent of a foreign government. See United States v. Verdugo-Urquidez, 856 F.2d 1214, 1216 (9th Cir. 1988), rev'd, 494 U.S. 259, 110 S.Ct. 1056, 108 L.Ed.2d 222 (1990) (Camarena murder suspect arrested by local Mexican police after U.S. arrest warrant was issued and suspect was handed over to U.S. Marshals at the U.S.-Mexico border); Matta-Ballesteros v. Henman, 896 F.2d 255, 256 (7th Cir. 1990) (Camarena murder suspect arrested in Honduras by Honduran Special Troops accompanied by U.S. Marshals; suspect driven to U.S. Air Force Base and flown to U.S.). Others were arrested in the United States. See United States v. Lopez-Alvarez, 970 F.2d 583, 586 (9th Cir. 1992); United States v. Felix-Gutierrez, 940 F.2d 1200, 1203 (9th Cir. 1991).

Despite the clear limitation on the extraterritorial reach of the arrest warrant, Sosa would have us believe that Alvarez's arrest in Mexico was authorized under American law.²⁴ The United States takes the same position in its defense against Alvarez's false arrest claim, which we discuss in a later section but which is also relevant here. Both parties conclude that the federal officers (and, by implication, Sosa) were authorized by statute to make warrantless arrests outside the United States. Because the criminal statutes under which Alvarez was charged have extraterritorial application, the argument goes, Congress must have granted DEA agents broad authority to enforce those statutes beyond our borders.

24 The district court emphasized that no warrant was issued by Mexican authorities and no Mexican official lawfully effectuated the arrest. Although the district court focused on this lack of local authority, our analysis centers on the DEA's authority under United States law. We do not hold that extraterritorial authority in this case rests on "the consent or assistance of the host country," despite the dissent's preoccupation with the subject in Section III.B. of its opinion.

The proper starting point is, of course, the applicable statutory scheme. We begin with a well-established canon of construction. "It is a longstanding principle of American law 'that legislation of Congress, unless a contrary intent appears, is meant to apply only within the territorial jurisdiction of the United States.'" EEOC v. Arabian Amer. Oil Co. ("Aramco"), 499 U.S. 244, 248, 111 S.Ct. 1227, 113 L.Ed.2d 274 (1991) (quoting Foley Bros. v. Filardo, 336 U.S. 281, 285, 69 S.Ct. 575, 93 L.Ed. 680 (1949)). "In applying this principle, 'we assume that Congress legislates against the backdrop of the presumption against extraterritoriality.'" Smith v. United States, 507 U.S. 197, 204, 113 S.Ct. 1178, 122 L.Ed.2d 548 (1993) (quoting Aramco, 499 U.S. at 248, 111 S.Ct. 1227). "[T]he presumption is rooted in a number of considerations, not the least of which is the commonsense notion that Congress generally legislates with domestic concerns in mind." Id. at 204 n. 5, 113 S.Ct. 1178. The canon also "serves to protect against unintended clashes between our laws and those of other nations which could result in international discord." Aramco, 499 U.S. at 248, 111 S.Ct. 1227 (citing McCulloch v. Sociedad Nacional de Marineros de Honduras, 372 U.S. 10, 20-22, 83 S.Ct. 671, 9 L.Ed.2d 547 (1963)).

The Supreme Court, in recognizing this principle, has carved out an exception for a narrow class of substantive criminal statutes. In United States v. Bowman, 260 U.S. 94, 43 S.Ct. 39, 67 L.Ed. 149 (1922), the Court reviewed a criminal fraud provision used to indict individuals who committed acts on a U.S. vessel outside of American territorial waters. The Court reiterated its presumption that, in most cases, if a substantive criminal provision is to be applied extraterritorially, "it is natural for Congress to say so in the statute." Id. at 98, 43 S.Ct. 39. But the Court found that "the same rule of interpretation should not be applied to criminal statutes which are, as a class, not logically dependent on their locality for the Government's jurisdiction, but are enacted because of the right of the Government to defend itself against obstruction, or fraud wherever perpetrated." Id.

We have no doubt that the substantive criminal statutes under which Alvarez was charged apply to acts occurring outside the United States. Invoking the rules of construction just described, we reasoned in United States v. Vasquez-Velasco, 15 F.3d 833, 839-41 (9th Cir. 1994), that 18 U.S.C. § 1959, the racketeering statute under which Alvarez was indicted, applied extraterritorially. Later, we applied the same principles to conclude that "Congress intended to apply statutes proscribing the kidnapping and murder of DEA agents extraterritorially." Felix-Gutierrez, 940 F.2d at 1204.

These cases reinforce the established proposition that certain criminal statutes are applicable to conduct occurring outside of the borders of the United States. It was precisely this principle of extraterritoriality that led the Supreme Court to conclude that Alvarez could be tried in the United States. Alvarez II, 504 U.S. at 657 n. 1, 112 S.Ct. 2188. And it is this same concept that is invoked in case after case to assert jurisdiction over defendants — whether United States or foreign nationals — for criminal conduct occurring outside of the United States. See, e.g., United States v. Neil, 312 F.3d 419, 421-23 (9th Cir. 2002) (applying extraterritoriality principle to bring citizen of St. Vincent and the Grenadines to trial in U.S. for sexual assault on cruise ship in Mexican territorial waters after cruise ship landed in U.S.); United States v. Hill, 279 F.3d 731, 739-40 (9th Cir. 2002) (applying harboring statute extraterritorially to bring to trial wife of violator of Deadbeat Parents Punishment Act arrested in U.S.); Chua Han Mow v. United States, 730 F.2d 1308, 1311-12 (9th Cir. 1984) (applying drug importation and distribution statutes extraterritorially to prosecute Malaysian defendant extradited to U.S.); Yousef, 327 F.3d at 87-111, 2003 U.S.App. LEXIS 6437, at **29-45 (applying provisions of the Destruction of Aircraft Act extraterritorially to conduct of terrorists who, after being arrested by Philippine and Malaysian police and later turned over to the FBI, were prosecuted for their participation in a conspiracy to bomb United States commercial airliners in Southeast Asia).²⁵

25 Congress has extended the United States' substantive criminal jurisdiction extraterritorially in a host of statutes, all of which state clearly their jurisdictional reach. See, e.g., 18 U.S.C. § 1119 (murder of U.S. national in a foreign country); 18 U.S.C. § 2332b (foreign terrorist activity in the U.S.); 18 U.S.C. §§ 1512(h), 1513(d) (witness tampering); 18 U.S.C. § 175 (use of biological weapons); 18 U.S.C. §§ 351, 1751 (crimes committed against high government officials); 18 U.S.C. § 1956 (money laundering); 18 U.S.C. § 2339B (assistance to foreign terrorist organizations); 18 U.S.C. § 1203(b)(1) (implementing Hostage Convention); 50 U.S.C. § 424 (extraterritorial jurisdiction over crimes relating to disclosure of national security information); 18 U.S.C. § 32(b) (violence against individual aboard or destruction of any "civil aircraft registered in a country other than the United States while such aircraft is in flight" or in service).

This proposition is not, however, the same as the far-reaching principle advocated by Sosa and the government, namely that a statute with extraterritorial application automatically carries with it the authority for United States agents to detain and arrest suspects worldwide. Extraterritorial application, in other words, does not automatically give rise to extraterritorial enforcement authority. Such a leap is too facile. That Congress may have intended the reach of a criminal statute to extend beyond our borders does not mean that Congress also intended to give federal law enforcement officers unlimited authority to violate the territorial sovereignty of any foreign nation to enforce those laws, or to breach international law in doing so. Bowman does not countenance such an extension, and our cases have never so held.²⁶

26 This basic distinction between the reach of the substantive criminal laws and the reach of law enforcement makes imminent sense in light of the myriad ways in which the United States regularly achieves lawful custody of persons located abroad. The options are many, ranging from purely formal means — such as extradition pursuant to a treaty or local statute, formal deportation, and revocation of passports — to purely diplomatic tactics, such as informal deportation and negotiation. See Abbell, supra, § 7-2, at 7-14-7-17.

In Bowman, the Supreme Court focused on the nature of the criminal conduct as a guide to determining the territorial reach of criminal statutes, but balanced that concern against limitations imposed by international law. The Court stated that "[t]he necessary locus, when not specially defined, depends upon the purpose of Congress as evinced by the description and nature of the crime and upon the territorial limitations upon the power and jurisdiction of a government to punish crime under the law of nations." 260 U.S. at 97-98, 43 S.Ct. 39. The Court repeatedly made reference to "the locus of the offense" and "the locus of [the] crime . . . in a foreign country," not to extraterritorial enforcement powers of the United States authorities. Id. at 97, 99, 43 S.Ct. 39. The court also emphasized that, by extending the reach of the substantive criminal statutes at issue, it was not imposing upon the sovereignty of other states.²⁷ Id. at 102-03, 43 S.Ct. 39.

27 The Court noted that because three of the defendants charged were citizens of the United States and were found in New York, "it is no offense to the dignity or right of sovereignty of Brazil to hold them for this crime." Bowman, 260 U.S. at 102, 43 S.Ct. 39. The Court expressly reserved the question whether the United States had jurisdiction over the fourth defendant, a citizen of Great Britain. Id. at 102-03, 43 S.Ct. 39.

Similarly, when we interpreted the criminal statutes for which Alvarez was indicted extraterritorially, we did so only with regard to the location of the conduct at issue. And even then we did so cautiously to ensure that we did not unnecessarily impinge on the sovereignty of other states or ignore accepted principles of international law. See Vasquez-Velasco, 15 F.3d at 839-40; Felix-Gutierrez, 940 F.2d at 1205-06; Chua Han Mow, 730 F.2d at 1311-12.

Taking the extraterritorial application of the applicable criminal laws as a given, the question then becomes whether Congress has separately authorized the unilateral, extraterritorial enforcement of those provisions in a foreign country by agents of the United States. The United States insists that such authority can be found in a provision in the Controlled Substances Act, 21 U.S.C. § 878, which grants certain powers to DEA and other law enforcement personnel.²⁸

28 Section 878 of the Act provides:

(a) Any officer or employee of the Drug Enforcement Administration or any State or local law enforcement officer designated by the Attorney General may —

(1) carry firearms;

(2) execute and serve search warrants, arrest warrants, administrative inspection warrants, subpoenas, and summonses issued under the authority of the United States;

(3) make arrests without warrant (A) for any offense against the United States committed in his presence, or (B) for any felony, cognizable under the laws of the United States, if he has probable cause to believe that the person to be arrested has committed or is committing a felony;

(4) make seizures of property pursuant to the provisions of this subchapter; and

(5) perform such other law enforcement duties as the Attorney General may designate.

21 U.S.C. § 878(a).

Subsection 878(a)(3) of that provision authorizes DEA agents to make warrantless arrests on probable cause for suspected felony violations. 21 U.S.C. § 878(a)(3). Although this subsection grants DEA agents felony arrest power, no language in the statute provides, or even suggests, that Congress intended that power to extend outside the borders of the United States. Given that the provision applies to DEA agents as well as "any State or local law enforcement officer designated by the Attorney General," it would in fact be anomalous to read subsection (3) as the statutory basis for a geographically limitless arrest power. Nor can such power be found in the catchall language of subsection (5), which states that DEA agents, as well as designated state and local officials, may "perform such other law enforcement duties as the Attorney General may designate." 21 U.S.C. § 878(a)(5). Again, nothing in the text of the statute remotely indicates that Congress sought to extend DEA arrest authority to any territory outside American borders.

Although legislative silence is not necessarily dispositive, these provisions must be construed against the backdrop of Aramco's presumption against extraterritoriality. Even the narrow Bowman exception offers no safe harbor.²⁹ Section 878(a) regulates executive authority, not criminal conduct. And this provision can hardly be classified as a "criminal statute which [is] . . . not logically dependent on [its] locality for the Government's jurisdiction." Bowman, 260 U.S. at 98, 43 S.Ct. 39. To hold otherwise would essentially swallow the presumption against extraterritoriality and grant, without express congressional authorization, worldwide law enforcement authority to United States officials (and to state and local officials upon designation by the Attorney General). Virtually a limitless number of statutes would have both extraterritorial reach and the prospect of extraterritorial enforcement. Surely such a result would all but eviscerate the longstanding principle that our laws generally apply only within our territorial borders.

29 We observe that Bowman's exception may be limited not only by its own language, but also in its application. Aramco did not mention Bowman at any point in its discussion of the presumption against extraterritoriality. We have interpreted the Court's silence as an indication that Bowman remains the law. See Felix-Gutierrez, 940 F.2d at 1205 n. 3. The Second Circuit, however, has held that Bowman should, at best, be interpreted narrowly. See Kollias v. D G Marine Maint., 29 F.3d 67, 71 (2d Cir. 1994) ("At best . . . the holding in Bowman should be read narrowly so as not to conflict with these more recent pronouncements on extraterritoriality."). Although we have implicitly rejected this latter interpretation, see, e.g., United States v. Corey, 232 F.3d 1166, 1170 (9th Cir. 2000), the Second Circuit's concerns underscore the fact that we should not cavalierly cast aside the presumption against extraterritoriality in the face of the Supreme Court's recent jurisprudence.

Faced with congressional silence on the matter, the United States analogizes this case to United States v. Chen, 2 F.3d 330 (9th Cir. 1993). The issue in Chen was whether agents of the Immigration and Naturalization Service acted outside their statutory authority by conducting an undercover investigation into the smuggling of Chinese aliens into the United States from international waters. The operation involved planting undercover agents on a chartered boat (the Corinthian) that rendezvoused with a Chinese ship some 320 miles off the coast of California. The agents watched and videotaped as the Chinese aliens boarded the Corinthian, keeping the aliens under surveillance during and after their entry into the United States. Id. at 332.

In evaluating whether the INS exceeded its statutory authority, we looked to 8 U.S.C. § 1103(a), the statute charging the Attorney General with enforcement of the Immigration and Nationality Act, and determined that Congress had given the Attorney General "extremely broad powers" to administer and enforce the immigration laws by directing the Attorney General to "perform such other acts as he deems necessary for carrying out his authority under the provisions of this chapter." Chen, 2 F.3d at 333 (citation and internal quotation marks omitted). We inferred from the broad language of § 1103(a) that "Congress intended to grant the Attorney General the corresponding power to enforce the immigration laws both within and without the borders of the United States." Id. We also pointed to § 1103(b), which specifically authorizes the Attorney General to delegate this broad authority to the Commissioner of the INS. Finally, we were careful to note that the Attorney General had in fact exercised this authority and had explicitly delegated her broad enforcement powers to the Commissioner under 8 C.F.R. § 2.1. Id. at 334. This chain of authority, we reasoned, provided "the legal basis for the INS and its agents to undertake offshore undercover investigations such as this one." Id.

But this case is not Chen. First, the INS operation in Chen, which consisted solely of observing and recording events, did not take place within the boundaries of another sovereign, but rather in international waters. That operation — unlike the abduction of a foreign citizen from a friendly neighbor — did not trigger any allegations of a breach of a law of nations. In fact, Chen did not even address international law, as traditional sovereignty concerns were not at issue. This distinction is critical, for one of the bedrock principles embodied in the presumption against extraterritoriality is that we must "protect against unintended clashes between our laws and those of other nations which could result in international discord." Aramco, 499 U.S. at 248, 111 S.Ct. 1227; see also Kollias, 29 F.3d at 70 (applying the same rationale). If Chen's expansion of INS authority to the high seas did not raise concerns about clashing with laws of another sovereign, the case before us most certainly presents that danger.

Second, the demonstrated chain of delegated authority on which Chen relied, extending from Congress to the Attorney General to the INS Commissioner to the INS agents, has not been shown to exist with respect to the DEA. Section 878(a)(3) does grant DEA agents broad authority to make warrantless arrests, and § 878(a)(5) does confer the authority to "perform such other law enforcement duties as the Attorney General may designate." 21 U.S.C. § 878(a)(5) (emphasis added). But even if Chen were to direct us to infer extraterritoriality from this bare language — a proposition that we do not accept — there is no evidence in this record that the Attorney General has in fact authorized the DEA Administrator to perform whatever extraterritorial enforcement powers the Attorney General may have — either generally or as to this abduction.³⁰

30 No regulation concerning the DEA's authority is analogous to the Attorney General's delegation of authority to the INS Commissioner in 8 C.F.R. § 2.1. In any event, there is no evidence that anyone ranking higher than the DEA Deputy Administrator or the United States Attorney for the Central District of California explicitly approved the operation. In view of this delegation vacuum, perhaps it is no surprise that the Department of Justice now requires explicit advance approval for such operations:

Due to the sensitivity of abducting defendants from a foreign country, prosecutors may not take steps to secure custody over persons outside the United States (by government agents or the use of private persons, like bounty hunters or private investigators) by means of Alvarez-Machain type renditions without advance approval by the Department of Justice. Prosecutors must notify the Office of International Affairs before they undertake any such operation. If a prosecutor anticipates the return of a defendant, with the cooperation of the sending State and by a means other than an Alvarez-Machain type rendition, and that the defendant may claim that his return was illegal, the prosecutor should consult with the OIA before such return.

Department of Justice, United States Attorneys' Manual, § 9-15.610.

The importance of obtaining specific authorization for extraterritorial law enforcement operations is brought into sharper relief by the fact that had the INS operation in Chen occurred within the boundaries of a foreign nation, rather than in international waters, the Attorney General (or the Commissioner, acting under delegated authority) would have been statutorily required to consult with the Secretary of State before deploying INS agents abroad. See 8 U.S.C. § 1103(a)(7) ("[A]fter consultation with the Secretary of State, [the Attorney General] may, whenever in his judgment such action may be necessary to accomplish the purposes of this chapter, detail employees of the Service for duty in foreign countries."). Such a restriction on the Attorney General's extraterritorial enforcement power, even in an area as obviously international as immigration, is evidence that Congress did not contemplate giving field agents the authority to act unilaterally in deciding to cross the borders of a friendly nation and abduct one of its citizens over that nation's objection. If the Attorney General must consult with the Secretary of State before dispatching INS agents to foreign lands, then surely, absent explicit statutory authorization, the Deputy Administrator of the DEA is not free to take it upon himself to send agents across the border into Mexico or to hire Mexican bounty hunters to act as surrogates to abduct a suspect.

Chen thus stands for only the proposition that the INS possesses limited delegated authority to conduct an operation on the high seas. At no point did we hold or even suggest that Congress has given license to the executive branch to violate international law in the course of enforcing criminal statutes that have extraterritorial reach. And surely Chen does not support the proposition that Congress has sub silencio delegated to the executive branch the authority to unilaterally enter a friendly nation and abduct one of its citizens in violation of international law.

Reading a generally worded statute like 21 U.S.C. § 878(a)(5) as evidence that Congress has given the DEA carte blanche to effectuate arrests within any sovereign state would require us to make the untenable assumption that Congress, in drafting such a statute, turned a blind eye to the interests of equal sovereigns and the potential violations of international law that would inevitably ensue.³¹ This we cannot do. See McCulloch, 372 U.S. at 21, 83 S.Ct. 671 (1963) ("'[A]n act of congress ought never to be construed to violate the law of nations if any other possible construction remains.'" (quoting Murray v. The Charming Betsy, 6 U.S. (2 Cranch) 64, 2 L.Ed. 208 (1804))).

31 Congress is well aware of the importance of respecting territorial sovereignty, and it has shown caution in expanding extraterritorial jurisdiction at the expense of this obligation. For instance, in passing the Omnibus Diplomatic Security and Anti-Terrorism Act of 1986, 22 U.S.C. § 4801 et seq., Congress refused to adopt a provision authorizing "self-help" measures. See Bills to Authorize Prosecution of Terrorists and Others Who Attack U.S. Government Employees and Citizens Abroad: Hearing on S. 1373, S. 1429, and S. 1508, Before the Subcomm. on Security and Terrorism of the Senate Comm. on the Judiciary, 99th Cong., 1st Sess. 63 (1985). Similarly, in passing the Anti-Drug Abuse Act of 1986, Congress required the Coast Guard to obtain foreign flag consent to board a foreign flag vessel on the high seas. Pub.L. No. 99-570, § 2015, 100 Stat. 3207, 3268 (repealed 1994).

We are not suggesting that Congress lacks the power to enact laws authorizing extraterritorial law enforcement powers. Nor do we question the powers of the political branches to override the principles of sovereignty in some circumstances, should the need arise. Rather, we are simply saying that we cannot impute such an intent where it is not expressed, and Congress has expressed no such intent here.³²

32 The dissent believes we should ignore well-established principles of statutory construction and give Congress the benefit of the doubt because we have recognized that "[d]elegation of foreign affairs authority is given . . . broader deference than in the domestic arena." Freedom to Travel Campaign v. Newcomb, 82 F.3d 1431, 1438 (9th Cir. 1996). But Freedom to Travel and the other nondelegation cases cited by the dissent are inapplicable here. We have no quarrel with the position that Congress, in giving the Executive authority over matters of foreign affairs, may delegate authority through broad (albeit not limitless) directives. See Zemel v. Rusk, 381 U.S. 1, 17, 85 S.Ct. 1271, 14 L.Ed.2d 179 (1965). Rather, we are simply saying that there is no evidence in the applicable statutory scheme that Congress ever granted the DEA the power to conduct arrests abroad. Hence, we do not address whether 21 U.S.C. § 878 is an impermissible delegation of congressional power.

Congress has shown that it is quite capable of making clear when arrest powers should have extraterritorial effect. See Aramco, 499 U.S. at 258, 111 S.Ct. 1227 ("Congress' awareness of the need to make a clear statement that a statute applies overseas is amply demonstrated by the numerous occasions on which it has expressly legislated the extraterritorial application of a statute."). In defining the law enforcement powers of the Coast Guard, for example, Congress provided that "[t]he Coast Guard may make . . . arrests upon the high seas and waters over which the United States has jurisdiction." 14 U.S.C. § 89(a). The powers of customs officials on the high seas have likewise been clearly articulated. See 19 U.S.C. § 1701 (permitting customs officials to seize or arrest in those areas of the high seas designated as customs-enforcement areas by the President).

More recently, in the Military Extraterritorial Jurisdiction Act of 2000,³³ Congress included clear and separate provisions pertaining both to the extraterritorial scope of the substantive crime and the executive agency's power to arrest. Section 3261(a), relating to certain members and employees of the Armed Forces, addresses the extraterritorial scope of the substantive crime:

33 This legislation was quickly enacted in response to the Second Circuit's decision in United States v. Gatlin, 216 F.3d 207 (2d Cir. 2000), which highlighted a gap in prosecutions of civilian personnel living abroad with the military.

Whoever engages in conduct outside the United States that would constitute an offense punishable by imprisonment for more than 1 year if the conduct had been engaged in within the special maritime and territorial jurisdiction of the United States . . . shall be punished as provided for that offense.

18 U.S.C. § 3261(a) (emphasis added). Section 3262(a), pertaining to "arrest and commitment," explicitly lays out the scope of arrest powers:

The Secretary of Defense may designate and authorize any person serving in a law enforcement position in the Department of Defense to arrest, in accordance with applicable international agreements, outside the United States any person described in section 3261(a) [of the Act] if there is probable cause to believe that such person violated section 3261(a).

18 U.S.C. § 3262(a) (emphasis added).³⁴ If Congress thought it could rely on courts to supply extraterritorial scope through searching interpretations of vague statutes, no such language would be necessary.

34 The government points to other statutes pertaining to the military's powers overseas, such as 10 U.S.C. § 374(b)(1)(D) and 18 U.S.C. § 351, arguing that these provisions "plainly envision foreign law enforcement activity." We agree. These statutes underscore the point that Congress is clear when it wishes to be. Section 374(b)(1)(D) allows the Secretary of Defense, upon the request of a federal law enforcement agency, to make defense personnel available "to operate equipment" with respect to "a rendition of a suspected terrorist from a foreign country to the United States to stand trial." Section 351 allows the FBI to request assistance from the military, as well as "any Federal, State, or local agency," in "investigat[ing]" kidnappings or assassinations of Congressional, Cabinet, and Supreme Court members. Not only do these statutes not speak to military arrest powers, but they define the universe (e.g., operating equipment or assisting in investigation) in which Congress has chosen to involve the military in law enforcement overseas. Section 374(b)(1)(D) is one of a number of provisions, along with the Posse Comitatus Act, 18 U.S.C. § 1385, that actually limit military involvement in civilian law enforcement operations. In considering 10 U.S.C. §§ 371-80, we concluded that "these sections impose limits on the use of American armed forces abroad." United States v. Khan, 35 F.3d 426, 431 n. 6 (9th Cir. 1994).

Wishful thinking is no substitute for clear congressional authority. Congress surely knows how and when to expand the reach of its laws beyond our borders. There is little doubt that Congress has the authority to do so; there is also little doubt that it has not done so here. Thus, although we recognize that the kidnapping and murder of DEA agents abroad necessitates the exercise of extraterritorial criminal jurisdiction, absent a clear directive, we cannot conclude that Congress has given the DEA unlimited enforcement powers abroad. Finding no basis in law for the DEA's actions, and left only with a warrant issued by a United States court, we conclude that Alvarez's arrest, and hence his detention, were arbitrary because they were not "pursuant to law." Consequently, Alvarez established a tort committed in violation of the law of nations.

II. ALIEN TORT CLAIMS ACT — SUBSTITUTION OF THE UNITED STATES FOR THE DEA AGENTS

We next consider whether the district court appropriately substituted the United States for the individual government defendants. The Federal Employees Liability Reform and Tort Compensation Act of 1988 (the "Westfall Act"), 28 U.S.C. § 2679, provides that, for civil actions arising out of the wrongful act of a federal employee acting within the scope of his official duties, the United States is to be substituted as a defendant and the claims may proceed only under the FTCA. 28 U.S.C. § 2679(b)(1). This exclusive remedy provision does not apply, however, in an action "which is brought for a violation of a statute of the United States under which such action against an individual is otherwise authorized." 28 U.S.C. § 2679(b)(2)(B). Alvarez argues that the ATCA falls within this exemption.

But we agree with the three-judge panel's conclusion that the exemption does not apply here, and that the United States was properly substituted for the individual DEA agents. Alvarez-Machain IV, 266 F.3d at 1053. Accordingly, we adopt the relevant portion of that opinion:

The district court held that an action under the ATCA was not exempt from the exclusive remedy provision of the Liability Reform Act. It reasoned that "it is international law, not the ATCA," that gives individuals fundamental rights. Therefore, a claim under the ATCA is based on a violation of international law, not of the ATCA itself.

This reading is consistent with the Supreme Court's reasoning in United States v. Smith, 499 U.S. 160, 111 S.Ct. 1180, 113 L.Ed.2d 134 (1991). In Smith, the Court rejected the argument that a claim for medical malpractice was "authorized" by the Gonzalez Act and therefore fit the 28 U.S.C. § 2679(b)(2)(B) exception for violations of a statute. The court explained: "[n]othing in the Gonzalez Act imposes any obligations or duties of care upon military physicians. Consequently, a physician allegedly committing malpractice under state or foreign law does not 'violate' the Gonzalez Act." Smith, 499 U.S. at 174, 111 S.Ct. 1180.³⁵ The same can be said of the ATCA. The language of § 1350 creates no obligations or duties. Admittedly, the ATCA differs from the Gonzalez Act in that it creates a cause of action for violations of international law, whereas the Gonzalez Act limited the common law liability of doctors. See Marcos II, 25 F.3d at 1475 (rejecting the argument that the ATCA is merely jurisdictional); Abebe-Jira v. Negewo, 72 F.3d 844, 848 (11th Cir. 1996); Filartiga, 630 F.2d at 885-86. Nonetheless, we find nothing in this distinction to cause us to deviate from the plain language of the statute. We therefore agree with the district court that Alvarez's claims under the ATCA were subject to substitution under the Liability Reform Act. Accordingly, Alvarez's exclusive remedy against the United States, in lieu of the DEA agents, is through the FTCA.

35 The relevant provision of the Gonzalez Act provides:

The remedy against the United States provided by[the FTCA] for damages for personal injury, including death, caused by the negligent or wrongful act or omission of any physician . . . of the armed forces . . . while acting within the scope of his duties or employment . . . shall hereafter be exclusive of any other civil action or proceeding by reason of the same subject matter against such physician . . . whose act or omission gave rise to such action or proceeding.

10 U.S.C. § 1089(a) (1994).

Id. at 1053-54.

Because the United States is substituted for the DEA agents, we treat the claims brought against the agents within the context of the FTCA. See § IV infra.

III. ALIEN TORT CLAIMS ACT — DAMAGES

A. CHOICE OF LAW

In addressing the matter of damages related to Sosa's liability under the ATCA, we must first determine the applicable substantive law. We review de novo the district court's decision concerning the appropriate choice of law. Abogados v. ATT, Inc., 223 F.3d 932, 934 (9th Cir. 2000).

Two obvious choices present themselves in this cross-border dispute: the domestic law of the United States and that of Mexico. The district court chose to apply federal common law, rather than Mexican law, in fashioning a damages award for Sosa's ATCA violations. The court reasoned that Mexican law would "inhibit the appropriate enforcement of the applicable international law or conflict with the public policy of the United States." Alvarez-Machain v. United States, No. 93-4072, slip op. at 33 (Sept. 9, 1999) (quoting Filartiga v. Pena-Irala ("Filartiga II"), 577 F.Supp. 860, 864 (E.D.N.Y. 1984)).

The precise issue before us, the choice of law for damages under the ATCA, is one of first impression. In Marcos III, we construed the district court's award of exemplary damages as having embraced Philippine law and concluded that this was not an error because such damages were allowed under Philippine law. 103 F.3d at 779-80. Our holding in Marcos III, however, went no further. We did not review the district court's choice of law analysis or enumerate the circumstances in which foreign law would apply. See id. (noting that there was "no ruling by the district court expressly choosing Philippine law").

The few courts that have addressed damages under the ATCA do not appear to have followed a consistent approach in determining the applicable law. Perhaps the most explicit treatment of the issue was offered by the district court in the Filartiga litigation. When faced with the question of damages on remand, the district court decided, in light of the ATCA's purpose, that federal choice of law principles should govern the initial determination of the remedy. See Filartiga II, 577 F.Supp. at 863. Applying these principles in the broadest of terms, the court noted that virtually all of the contacts took place in Paraguay, and thus Paraguayan law appeared to be appropriate for setting compensatory damages. Id. at 863-64. The court took a different tack, however, on punitive damages. Because Paraguay did not recognize punitive damages, which were deemed necessary "to give effect to the manifest objectives of the international prohibition against torture," the court turned to international law principles. Id. at 865.

Other courts awarding damages in the wake of Filartiga II have adopted a number of approaches. Most courts have not directly addressed the choice of law dilemma, while others have offered variations on the Filartiga II theme. See, e.g., Tachiona v. Mugabe, 234 F.Supp.2d 401, 418-22 (S.D.N.Y. 2002) (addressing the choice of law issue, but abandoning a traditional choice of law analysis in favor of a more "flexible" approach for determining both substantive rights and remedies); Mehinovic v. Vuckovic, 198 F.Supp.2d 1322, 1358-59 (N.D.Ga. 2002) (conducting no choice of law analysis but making repeated references to "international law" in awarding both compensatory and punitive damages); Xuncax, 886 F.Supp. at 183, 198 (using an analysis similar to that of Tachiona); Avril, 901 F.Supp. at 335 (citing Filartiga II for the position that both compensatory and punitive damages are available but providing no indication as to which law was applied).³⁶

36 It bears noting that most of the cases addressing damages under the ATCA have done so without the benefit of, or without reference to, Zicherman v. Korean Air Lines Co., 516 U.S. 217, 229, 116 S.Ct. 629, 133 L.Ed.2d 596 (1996), in which the Supreme Court interpreted the damages provisions of the Warsaw Convention and concluded that it does not "empower us to develop some common-law rule — under cover of general admiralty law or otherwise — that will supersede the normal federal disposition." The Court held that the Convention "provide[d] nothing more than a pass-through, authorizing us to apply the law that would govern in the absence of the Warsaw Convention," which in that case was the Death on the High Seas Act, 46 U.S.C.App. § 761. Zicherman, 516 U.S. at 229, 116 S.Ct. 629.

Mindful of this varied landscape, we begin our inquiry with a traditional choice of law analysis. As the Supreme Court has counseled, "[c]hoice of law is, of course, determined by the forum jurisdiction," Zicherman, 516 U.S. at 228-29, 116 S.Ct. 629, which in this case is federal court. Federal question jurisdiction was predicated on the ATCA and thus federal common law applies to the choice-of-law determination. See Chan v. Soc'y Expeditions, Inc., 123 F.3d 1287, 1297 (9th Cir. 1997) (holding that federal common law applies to choice-of-law determination in federal question case).³⁷ Under federal common law, we look to the Restatement (Second) of Conflict of Laws ("Restatement of Conflicts") for guidance. Schoenberg v. Exportadora de Sal, S.A., 930 F.2d 777, 782 (9th Cir. 1991) (explaining, in the context of the Foreign Sovereign Immunities Act, that "[f]ederal common law follows the approach of the Restatement (Second) of Conflict of Laws"); see also Bickel v. Korean Air Lines Co., 83 F.3d 127, 130 (6th Cir. 1996) (noting, in the context of the Warsaw Convention, that "[i]n the absence of any established body of federal choice of law rules, we begin with the Restatement (Second) of Conflict of Laws. . . .").

37 Although the Second Circuit observed in Pescatore v. Pan American World Airways, Inc., 97 F.3d 1, 12 (2d Cir. 1996), that "the law is unsettled when it comes to applying either a federal common law choice of law rule or state choice of law principles in nondiversity cases," we believe that both Zicherman and our precedent support the application of federal common law conflicts principles.

Section 145³⁸ of the Restatement, which delineates the general principles applicable to torts, states that the "rights and liabilities of the parties with respect to an issue in tort are determined by the local law of the state which, with respect to that issue, has the most significant relationship to the occurrence and the parties under the principles stated in § 6."³⁹ The section continues by listing the following "contacts" that should "be taken into account in applying the principles of § 6 to determine" the state with the "most significant relationship":

38 Sosa urges us to look to § 146 of the Restatement which provides, with respect to personal injury actions, that there is a presumption in favor of applying "the local law of the state where the injury occurred," which in this case was Mexico. But the tort here — arbitrary arrest and detention as a recognized violation of international law — is not a classic personal injury claim. Nor does Alvarez's claim "involve either physical harm or mental disturbance . . . resulting from physical harm" as envisioned by § 146. Restatement of Conflicts § 146 cmt. b. Finally, the presumption is not absolute and other considerations weigh in favor of applying United States law.

39 The factors in § 6 include:

(a) the needs of the interstate and international systems,

(b) the relevant policies of the forum,

(c) the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue,

(d) the protection of justified expectations,

(e) the basic policies underlying the particular field of law,

(f) certainty, predictability and uniformity of result, and

(g) ease in the determination and application of the law to be applied.

Id. § 6(2).

(a) the place where the injury occurred,

(b) the place where the conduct causing the injury occurred,

(c) the domicil, residence, nationality, place of incorporation and place of business of the parties, and

(d) the place where the relationship, if any, between the parties is centered.

Id. § 145(2).

These principles are meant to serve as a guide for consideration of competing policy choices. The factors, coupled with the contacts, are not necessarily of equal weight, nor do they lend themselves to a bean-counting exercise in which everything is lined up on a ledger and the answer emerges. Indeed, as noted in the comment, "[a]t least some of the factors . . . will point in different directions in all but the simplest case." Id. § 6 cmt. c. This international dispute illustrates in practical terms the reality of that admonition. In a claim based on a universal, international standard, it may seem presumptuous to choose the law of one country over another. Admittedly, the needs of the international system are often too complex to dictate a clear choice, in part because our task is limited to a legal analysis and we leave foreign policy to the Executive branch and the diplomats. Nonetheless, we are driven to make a choice so that damages may be assessed in accord with the substantive law of a chosen jurisdiction.

Stepping back and looking at the overall picture, we view this case as a series of events that began and ended in the United States, and which are inextricably intertwined with the United States government. The United States' interests are particularly pointed here: the United States itself is a party, and it is the conduct of the United States government, in its efforts to bring a suspect to justice, that spawned the international incident. The genesis of the crucial events was a federal criminal prosecution of Alvarez in Los Angeles. DEA agents working in the United States devised a plan, which they hired Sosa to carry out, and without which the tort would not have occurred. Sosa acted according to DEA instructions when he helped detain Alvarez and transport him to the United States for trial. Sosa himself had no justifiable expectation that Mexican law would apply, particularly because he was employed as an agent of the American government, and because this is a tort, rather than a contract, case. The relationship between Sosa and Alvarez was intimately connected with, and a direct product of, the interests of the United States government. Just as importantly, the tort is predicated on an arrest and detention that were arbitrary because the agents exceeded the scope of their authority under United States law.

As Sosa points out, some of the Restatement factors weigh in favor of applying Mexican law. Alvarez's actual arrest occurred in Mexico. Both Alvarez and Sosa were Mexican citizens and residents at the time of the events in question (although Sosa later moved to the United States). As a result, Mexico may in fact have competing interests — seeking to obtain compensation for its citizen, Alvarez, while limiting damages from Sosa, another of its citizens.

Nonetheless, we must also take into account the policy of the United States, as expressed in the ATCA, to provide a remedy for violations of the law of nations. See Marcos II, 25 F.3d at 1475. We agree with the district court that limitations on damages under Mexican law — including the unavailability of punitive damages — are not consistent with the congressional policy that underlies the ATCA.

After weighing these factors, we conclude that the relative importance of United States contacts and interests counsels in favor of applying United States law. Our ruling today does not foreclose the application of foreign law in another circumstance; it is simply the appropriate outcome given the factors and policies present in this suit.

Our choice of law conclusion brings us to another level of inquiry: In applying United States law, should we apply federal common law or the law of California? We are aware of the Supreme Court's view that we should not reach out to extend federal common law. See O'Melveny Myers v. FDIC, 512 U.S. 79, 83-84, 114 S.Ct. 2048, 129 L.Ed.2d 67 (1994); see also Kamen v. Kemper Fin. Servs., Inc., 500 U.S. 90, 98, 111 S.Ct. 1711, 114 L.Ed.2d 152 (1991) (explaining the presumption in favor of incorporating state law to provide the content of federal common law, and that "a court should endeavor to fill the interstices of federal remedial schemes with uniform federal rules only when the scheme in question evidences a distinct need for nationwide legal standards . . . or when express provisions in analogous statutory schemes embody congressional policy choices readily applicable to the matter at hand. . . . .").

On the other hand, because the ATCA invokes international law principles of universal concern, it holds a unique place among federal statutory tort causes of action, and application of federal common law is therefore appropriate.⁴⁰ See Texas Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 641, 101 S.Ct. 2061, 68 L.Ed.2d 500 (1981) (observing that, in "international disputes implicating . . . relations with foreign nations . . . our federal system does not permit the controversy to be resolved under state law" because the "international nature of the controversy makes it inappropriate for state law to control"); see also Sabbatino, 376 U.S. 398, 427 n. 25, 84 S.Ct. 923, 11 L.Ed.2d 804 (1964) (noting that the ATCA is an example of a statute reflecting a "concern for uniformity in this country's dealings with foreign nations").

40 Although we apply federal common law, we note that, as discussed below, the result would be the same under state law.

B. SCOPE OF DAMAGES

The district court determined that Alvarez could recover damages only for activities taking place prior to the point that United States law enforcement authorities took him into custody, not for the entire period in which he was imprisoned in the United States. We review this question of law de novo. See United States v. Stephens, 237 F.3d 1031, 1033 (9th Cir. 2001).

There is no established body of case law applying federal common law to determine the proper scope of damages for arbitrary arrest and detention. Although several federal cases have awarded damages for this brand of international law violation, none of those cases dealt with the unique set of facts presented here. See, e.g., Xuncax, 886 F.Supp. at 197-98 (awarding damages for arbitrary detention authorized by Guatemala's Minister of Defense). Even so, we agree with the district court that existing principles governing false arrest provide adequate guidance.

In the context of law enforcement, the federal courts are largely in accord that, consistent with the principles of tort law, the chain of causation set in motion by the initial act of misconduct of one actor can be broken by the acts of a third party. For example, police officers have been held to be insulated from liability for deprivations of liberty where there are independent, intervening acts of other decisionmakers in the criminal justice system, such as prosecutors, grand juries, or judges. See Heck v. Humphrey, 512 U.S. 477, 484, 114 S.Ct. 2364, 129 L.Ed.2d 383 (1994) ("'If there is a false arrest claim, damages for that claim cover the time of detention up until issuance of process or arraignment, but not more.'" (quoting W. Keeton, D. Dobbs, R. Keeton, D. Owen, Prosser Keeton on the Law of Torts 888 (5th ed. 1984))); Townes v. City of New York, 176 F.3d 138, 147 (2d Cir. 1999) (holding that the trial judge's independent decision not to suppress evidence, though erroneous, broke the chain of causation for purposes of police officer's liability); Barts v. Joyner, 865 F.2d 1187, 1195 (11th Cir. 1989) (holding that intervening acts of prosecutor, grand jury, and judge broke chain of causation); Hand v. Gary, 838 F.2d 1420, 1427-28 (5th Cir. 1988) (holding that a sheriff's actions were not the proximate cause of damages given intervening acts of federal agents, federal prosecutors, and grand jury). In this connection, we have held that the "[f]iling of a criminal complaint immunizes investigating officers . . . from damages suffered thereafter because it is presumed that the prosecutor filing the complaint exercised independent judgment in determining that probable cause for an accused's arrest exists at that time." Smiddy v. Varney, 665 F.2d 261, 266 (9th Cir. 1981).⁴¹

41 Our holding in Smiddy was limited. We concluded that the presumption that the prosecutor exercised independent judgment can be rebutted by, for instance, "a showing that the [prosecutor] was pressured or caused by the investigating officers to act contrary to his independent judgment," or by "the presentation by the officers to the [prosecutor] of information known by them to be false." 665 F.2d at 266-67.

These principles of proximate causation, taken in combination with the Supreme Court's holding in Alvarez-Machain II, guide us in assessing the scope of Sosa's liability. Sosa's participation in Alvarez's arrest and detention in this case took place almost solely within the confines of Mexico. Although he was guided by the unlawful directives of American DEA agents, once he delivered Alvarez to United States authorities in El Paso, the actions of domestic law enforcement set in motion a supervening prosecutorial mechanism which met all of the procedural requisites of federal due process and ultimately received the blessing of the United States Supreme Court. See Alvarez-Machain II, 504 U.S. at 669-70, 112 S.Ct. 2188. To be sure, a grand jury had already indicted Alvarez and an American arrest warrant had been issued by the time Sosa was hired, giving this case a unique factual twist when compared to traditional false arrest cases. But, as we have explained, these procedural formalities stand apart from the illegitimacy that characterized Alvarez's initial arrest and detention, and came into operation only at the moment Alvarez set foot on U.S. soil. At that point, the criminal justice system, with proper jurisdiction, began its march toward trial and the chain of causation linked to Sosa's actions was broken, thus limiting Sosa's liability for damages.

Because the district court cited California law for its damages analysis, the parties focus on the nuances of California law, despite framing the issue in terms of choosing either federal common law or Mexican law. Applying California law, however, yields the same result. The California Supreme Court recently rejected a plaintiff's claim of false imprisonment for the entire time he was held in custody. See Asgari v. City of Los Angeles, 15 Cal.4th 744, 63 Cal.Rptr.2d 842, 937 P.2d 273, 281 (1997). Relying on state statutes, the court explicitly distinguished the rebuttable presumption rule of Smiddy, clarifying that a police officer's liability for false arrest could not, even with a showing of bad faith, include damages caused by incarceration following arraignment because that result would thwart the applicable statutes' directives. Id. at 279.

Alvarez seeks to distinguish Asgari by arguing that it is a narrow holding based only on immunity principles grounded in the California Tort Claims Act, specifically Cal. Gov. Code §§ 820.4 and 821.6. Instead, he urges us to rely on an earlier case, Gill v. Epstein, 62 Cal.2d 611, 44 Cal.Rptr. 45, 401 P.2d 397 (1965), which held, prior to the enactment of the above provisions, that a plaintiff could recover damages arising from his incarceration after his arraignment because the arraignment was not an independent act that could break the chain of causation. Id. at 401. But Gill does not help Alvarez. In Gill, the plaintiff was arrested without a warrant, and the case was dismissed at a preliminary hearing five days after the arraignment. Id. at 398-99. The court held that the plaintiff could recover for damages up until the time an independent judgment was made as to probable cause for his arrest. See id. at 401. Here, there is no question that at the time Alvarez was arrested, an independent judgment had already been made that he should be brought to trial.⁴² As a result, Alvarez is entitled to damages only to the point at which he was handed over to U.S. authorities.

42 Although we decline to speak for the California Supreme Court as to the status of Gill after Asgari, we note also that in Asgari the court looked not only to statutory immunity principles but also to the broader proximate cause principles articulated in New York's Broughton rule, which measures liability only up to the time of arraignment or indictment, whichever comes first. See Asgari, 63 Cal.Rptr.2d 842, 937 P.2d at 281 n. 10 (citing Broughton v. State, 37 N.Y.2d 451, 373 N.Y.S.2d 87, 335 N.E.2d 310, 316 (1975)).

IV. FEDERAL TORT CLAIMS ACT

The FTCA acts as a limited waiver of the sovereign immunity of the United States for certain torts committed by its employees. 28 U.S.C. §§ 1346(b), 2674. The statute provides that the United States shall be "liable . . . in the same manner and to the same extent as a private individual under like circumstances." 28 U.S.C. § 2674. At issue here is whether Alvarez's claims fit within the FTCA's waiver provision or instead fall specifically within any of the statutory exclusions to FTCA jurisdiction — in particular, the "foreign activities" exception or the "intentional tort" exception.

The United States argues that Alvarez's kidnapping lies outside the jurisdiction of the FTCA. But we agree with the district court that neither exception applies.

A. "FOREIGN ACTIVITIES" EXCEPTION

The foreign activities exception bars recovery for "[a]ny claim arising in a foreign country." 28 U.S.C. § 2680(k). Its purpose is "to ensure that the United States is not exposed to excessive liability under the laws of a foreign country over which it has no control." Nurse v. United States, 226 F.3d 996, 1003 (9th Cir. 2000). The district court held that many of Alvarez's claims, such as assault and the resulting infliction of emotional distress, derived from acts that took place entirely in Mexico and so were excluded under the ATCA. Alvarez does not appeal that decision.

But the district court permitted other claims — false arrest, false imprisonment, and the resulting infliction of emotional distress — to go forward under the "headquarters doctrine." Because "[t]he entire scheme of the FTCA focuses on the place where the negligent or wrongful act or omission of the government employee occurred," Sami v. United States, 617 F.2d 755, 761 (D.C. Cir. 1979), a claim can still proceed under the headquarters doctrine if harm occurring in a foreign country was proximately caused by acts in the United States. See Nurse, 226 F.3d at 1003; see also Cominotto v. United States, 802 F.2d 1127, 1130 (9th Cir. 1986) (holding that an FTCA claim arises where an act or omission occurs and "not necessarily at the site of the injury or the place where the negligence has its operative effect" (internal quotation marks omitted)).

The quintessential headquarters claim involves federal employees working from offices in the United States to guide and supervise actions in other countries. See Nurse, 226 F.3d at 1003 (applying the doctrine to FTCA claims made by a Canadian detained in Vancouver, British Columbia, against the U.S.-based Customs officials who trained the Vancouver agents); Couzado v. United States, 105 F.3d 1389, 1395-96 (11th Cir. 1997) (applying the doctrine to claims against DEA agents in the United States who coordinated an arrest in Honduras); Sami, 617 F.2d at 761-63 (applying the doctrine to claims against the Chief of the United States National Central Bureau in Washington, D.C., who sent messages causing an improper arrest in Germany). In evaluating whether the headquarters doctrine applies, we look to the law of the state where the alleged act occurred — in this case, California. See 28 U.S.C. § 1346(b)(1); Couzado, 105 F.3d at 1395 (applying Florida law to determine whether the doctrine applies to alleged negligence by DEA officials who were based in Florida and caused harm in Honduras).

Alvarez's abduction fits the headquarters doctrine like a glove. Working out of DEA offices in Los Angeles, Berellez and his superiors made the decision to kidnap Alvarez and, through Garate, gave Barragan precise instructions on whom to recruit, how to seize Alvarez, and how he should be treated during the trip to the United States. DEA officials in Washington, D.C., approved the details of the operation. After Alvarez was abducted according to plan, DEA agents supervised his transportation into the United States, telling the arrest team where to land the plane and obtaining clearance in El Paso for landing. The United States, and California in particular, served as command central for the operation carried out in Mexico.

By contrast, we see little resemblance to the facts of Cominotto, in which we rejected the DEA informant's headquarters claim because he had disobeyed Secret Service orders by jumping into the suspects' car late one night in Bangkok. 802 F.2d at 1130. Alvarez did little but serve as an unsuspecting target of an operation planned in the United States. Under California law, negligent or criminal acts carried out by Alvarez's abductors in furtherance of the objectives given to them by American DEA agents "do not break the causal link between" the conduct of the DEA agents and Alvarez's injuries. Vickers v. United States, 228 F.3d 944, 956 (9th Cir. 2000). The arrest team's seizure of Alvarez was not the interruption, but the fulfillment, of the DEA agents' tortious acts. The events for which Alvarez seeks relief occurred precisely as the DEA intended.

The United States offers little to support its alternative argument that, even if applicable, the headquarters doctrine does not apply to intentional torts. We see no valid reason to distinguish between negligence and intentional torts when the purpose of the doctrine is to hold the federal government responsible where the plaintiff's injuries are proximately caused by conduct in the United States. Sami, 617 F.2d at 762 (noting that examination of the legislative history shows that the foreign activities exception "does not apply if the wrongful acts or omissions complained of occur in the United States" (emphasis added)). We hold that the headquarters doctrine applies to both negligence and intentional torts. Alvarez's kidnapping claim therefore does not fall within the foreign activities exception.

B. "INTENTIONAL TORT" EXCEPTION

We also agree with the district court that Alvarez's claims do not fall within the "intentional tort" exception to the FTCA. See 28 U.S.C. § 2680(h).⁴³ Although the waiver of sovereign immunity under the FTCA excludes intentional torts such as false arrest, this exclusion is followed by an important proviso: It does not apply if the intentional tort is committed by an "investigative or law enforcement officer." Id. See also Orsay v. United States Dep't of Justice, 289 F.3d 1125, 1134 (9th Cir. 2002) (noting that Congress chose "to single out investigative and law enforcement officers from other federal employees" because their "authority to use force and threaten government action carries with it the risk of abuse, or the risk of intentionally tortious conduct").

43 The Act provides an exception for

(h) Any claim arising out of assault, battery, false imprisonment, false arrest, malicious prosecution, abuse of process, libel, slander, misrepresentation, deceit, or interference with contract rights: Provided, That, with regard to acts or omissions of investigative or law enforcement officers of the United States Government, the provisions of this chapter and section 1346(b) of this title shall apply to any claim arising, on or after the date of the enactment of this proviso, out of assault, battery, false imprisonment, false arrest, abuse of process, or malicious prosecution. For the purpose of this subsection, "investigative or law enforcement officer" means any officer of the United States who is empowered by law to execute searches, to seize evidence, or to make arrests for violations of Federal law.

The DEA agents who orchestrated Alvarez's arrest are law enforcement officers as defined by the FTCA because they are "empowered by law to execute searches, to seize evidence, or to make arrests for violations of Federal law." 28 U.S.C. § 2680(h). Because the primary tortious act was the initiation and planning of Alvarez's abduction by the DEA agents, his claim falls squarely within this law enforcement proviso, and thus the intentional tort exclusion does not apply.

The purpose of the law enforcement proviso in § 2680(h) is to "provid[e] a remedy against the Federal Government for innocent victims of Federal law enforcement abuses." Orsay, 289 F.3d at 1134-35 (quoting S.Rep. No. 93-588, 93d Cong., 2d Sess. 3 (1973), reprinted in 1974 U.S. Code Cong. Admin. News 2789, 2792 (1974)). As the original three-judge panel put it, this purpose would be manifestly frustrated if law enforcement officers could avoid liability by recruiting civilians "to do their dirty work." Alvarez-Machain IV, 266 F.3d at 1056.

Because neither the "foreign activities" exception nor the "intentional tort" exception applies, we proceed to the merits of Alvarez's false arrest claim under the FTCA.

C. FALSE ARREST CLAIM

The parties agree that if no exception applies, California law determines whether and to what extent the United States is liable. See 28 U.S.C. § 1346(b)(1). Because "[u]nder California law, a California court would apply federal law to determine whether an arrest by a federal officer was legally justified and hence privileged," the United States' liability hinges on whether federal employees "complied with applicable federal standards" in seizing Alvarez. Rhoden v. United States, 55 F.3d 428, 431 (9th Cir. 1995) (per curiam). The government argues that there is no California false arrest because federal law authorized Alvarez's apprehension in Mexico. Our earlier discussion of liability under the ATCA applies with equal force to our analysis of the FTCA claims against the United States. The DEA agents had no authority under federal law to execute an extraterritorial arrest of a suspect indicted in federal court in Los Angeles. See supra at § I.B.2.

Notwithstanding the fact that California law looks to federal law to determine the lawfulness of an arrest by federal officers, the district court concluded that Alvarez's abduction could still be justified as a citizen arrest under California law. The United States urges us to reach the same conclusion, arguing that, in certain situations, California's citizen arrest provision authorizes federal agents to make arrests even where federal authority is lacking.⁴⁴ See United States v. DeCatur, 430 F.2d 365, 367 (9th Cir. 1970) (noting that the arrest of the plaintiff by federal postal agents would have been justified under California Penal Code § 837, even if the agents lacked authority under a federal statute); People v. Crusilla, 77 Cal. App.4th 141, 91 Cal.Rptr.2d 415, 421 (1999) (holding that a federal immigration inspector's arrest of defendant was authorized as a citizen arrest).

44 California Penal Code § 837 permits a private citizen to arrest a person "[w]hen a felony has been in fact committed, and he has a reasonable cause for believing the person arrested to have committed it."

Reliance on the California law of citizen arrest is misplaced in this context. Although the FTCA holds the United States liable in the same way that a private person would be liable "under like circumstances," 28 U.S.C. § 2674, the law enforcement obligations and privileges of the DEA agents "make the law of citizen arrests an inappropriate instrument for determining FTCA liability." Arnsberg v. United States, 757 F.2d 971, 979 (9th Cir. 1985); see also 21 U.S.C. § 878(a)(2); Ting v. United States, 927 F.2d 1504, 1514 (9th Cir. 1991) (citing Arnsberg). In Arnsberg, we declined to require Internal Revenue Service agents, who arrested the plaintiff with a defective warrant, to meet the stricter standard for citizen arrests under Oregon law. 757 F.2d at 978-79. Instead, we concluded, "[t]he proper source for determining the government's liability" is "the law governing arrests pursuant to warrants." Id. at 979. Applying that law, we determined that the agents acted properly. Id.

The principle adopted in Arnsberg works both ways: just as the law of citizen arrest cannot be used to limit the authority of law enforcement officers, nor can it be used to extend that authority, by proxy, beyond its territorial limits.⁴⁵ The DEA agents, not the Mexican nationals, identified Alvarez and planned the operation in detail; Alvarez's abductors acted merely as pawns. In this situation, the law of citizen arrest simply does not apply.

45 The district court noted that Cal.Penal Code § 837 probably permits non-Californians to make a citizen arrest in California. California courts have applied this provision to police officers who make arrests outside of their jurisdiction but within California. See, e.g., People v. Monson, 28 Cal.App.3d 935, 105 Cal.Rptr. 92, 95 (1972). California law also permits Mexican police crossing the border in fresh pursuit of a suspect to make an arrest in California, provided that the official brings the prisoner before a magistrate in the county where the arrest occurred. See Cal.Penal Code § 852.2. However, these provisions do not authorize a planned, transborder abduction of an alien by either law enforcement authorities or private citizens.

Although, as in Arnsberg, we apply the law governing arrests pursuant to warrants, we see a world of difference between the acts of the law enforcement officers in Arnsberg and the DEA agents who planned Alvarez's abduction. In Arnsberg, the IRS officials acted "nearly perfectly," consulting with the United States Attorney and arresting the plaintiff pursuant to a warrant with only a minor discrepancy. Arnsberg, 757 F.2d at 979. In contrast, as we have discussed, the DEA agents here had no authority, statutory or otherwise, to effect an extraterritorial arrest. Nor did their minions across the border, who could no more claim a lawful privilege to arrest Alvarez than could the DEA agents themselves under the same circumstances. The district court that issued Alvarez's arrest warrant had no jurisdiction to issue a warrant for an arrest in Mexico. See Fed.R.Crim.P. 4(c)(2). Accordingly, the DEA agents authorized a false arrest against Alvarez. We reverse the district court's dismissal of the FTCA claims and remand for further proceedings.

CONCLUSION

In summary, we affirm the judgment with respect to Sosa's liability under the ATCA, albeit on different grounds than the district court. We also affirm the substitution of the United States for the DEA agents, the choice of United States rather than Mexican law to determine damages, and the limitation of damages to Alvarez's time in captivity in Mexico. We reverse and remand the district court's dismissal of the FTCA claims against the United States. We approve the dismissal of Garate. Each party shall bear its own costs on appeal.

AFFIRMED IN PART, REVERSED IN PART, AND REMANDED.

FISHER, Circuit Judge, with whom Chief Judge SCHROEDER and Circuit Judges GOODWIN, THOMAS, and PAEZ join, concurring:

I fully concur in the majority opinion, but write separately to articulate another ground on which I base my conclusion that Alvarez's arrest and detention were arbitrary because they were conducted without lawful authority. As the majority opinion explains, whatever power the political branches might have to override the principle of territorial sovereignty, Congress has not expressed its intent to delegate that power to the Drug Enforcement Agency. I would add, moreover, that to the extent the Executive branch has the power to act without congressional sanction, there has been no showing that that power was properly invoked here.

It may well be, as the majority and the dissenting opinions assume, that the Executive — like Congress — has the authority to breach another nation's sovereignty and override other norms of international law, should the need arise. It is evident, however, that neither Congress nor the Executive has expressed an intent to allow sub-Cabinet-level law enforcement officials in the DEA to be the final arbiters of that authority. The President, the Attorney General, the Secretary of State, perhaps the Secretary of Defense, the National Security Advisor — these are the proper Executive branch officials with whom to entrust the weighty decision to kidnap and arrest a suspect on friendly foreign soil. As Judge Gould acknowledges in his dissent, "the capture of a foreign national on foreign soil is no ordinary law enforcement choice; rather, it is a serious foreign policy decision," as evidenced by the international outcry that occurred in the wake of Alvarez's abduction. Such a decision, involving political judgments and national risks of the highest order, is — with all due respect to those involved — well above the pay-grade of those who approved the abduction of Alvarez here.

DEA Agent Hector Berrellez made the offer to pay Mexican nationals to apprehend Alvarez in Mexico and deliver him to the United States. Berrellez received authorization to make this offer from his superiors in the Los Angeles office of the DEA and from DEA Deputy Administrator Pete Gruden in Washington, D.C. There is no evidence that anyone ranking higher than the DEA Deputy Administrator or the United States Attorney for the Central District of California explicitly approved the operation.¹

1 There are some indications that individuals in the Attorney General's office were informed of the plan. At an evidentiary hearing before the district court on Alvarez's motion to dismiss the indictment in his criminal case, Berrellez testified that he believed that the Attorney General's office had been "consulted" about the operation. See United States v. Caro-Quintero, 745 F.Supp. 599, 603 (C.D.Cal. 1990). Berrellez did not, however, explain the basis for this belief; nor did he claim that the Attorney General's office gave its approval for the operation upon consulting with the DEA.
The only other evidence suggesting that the Attorney General's office might have been informed of the operation is an anonymous memorandum entitled "Operation Leyenda: Chronology" that the United States produced during discovery. According to the memorandum, whose origin is unclear, the United States Attorney's Office in Los Angeles approved the kidnapping plan, the DEA Administrator was "advised of the general plan, and he in turn advise[d] the Executive Assistant to the Attorney General." However, DEA Administrator John C. Lawn testified at his deposition that he had no advance knowledge of the plan to use Mexican nationals to apprehend Alvarez, and the United States denied Alvarez's request for admission that Lawn approved the operation. The assertions contained in Berrellez's testimony and in the memorandum are therefore unsupported by the evidence. Even if we were to accept the assertions as true, however, they indicate nothing more than that the Attorney General's office had knowledge of the operation. They do not support the further conclusion that the Attorney General explicitly sanctioned Alvarez's extraterritorial abduction.

It seems obvious that such a controversial, risky operation should have been evaluated and approved by the Attorney General personally (or at least by some highranking Department of Justice official authorized to act on the Attorney General's behalf), probably in consultation with the Department of State or the White House Counsel's office. Indeed, Judge Gould's dissent is predicated on the notion that "extraordinary renditions" of suspects from foreign nations are such critical foreign policy decisions that they must be planned and coordinated at the highest levels of government. He would rob this notion of any force, however, by equating the DEA actors here with the upper echelon of the Executive branch.

The Chen case illustrates that the decision to engage in extraterritorial operations is beyond the discretion of law enforcement officers. In stark contrast to the abduction of Alvarez, the limited INS operation at issue in Chen — as the Chen opinion took pains to delineate — was carefully planned and subjected to high-level review and approval at various levels within the Department of Justice:

On August 21, 1991, the Undercover Operations Review Committee of the United States Department of Justice (Review Committee) authorized the INS agents to proceed with the proposed undercover operation involving the use of the Corinthian in international waters.

. . . .

On August 27, the Review Committee considered a revised plan for the INS's proposed undercover operation. . . . The Review Committee was aware that INS agents would conduct the undercover investigation in international waters when approval was given. The Review Committee guidelines specifically contemplate an INS "undercover operation [that] will be conducted substantially outside the United States." INS Undercover Operation Guidelines at IV.A.(2). The operation was also approved by the United States Attorney for the Central District of California and an Assistant Attorney General in the Department of Justice, both of whom knew that the operation would be conducted in international waters.

Chen, 2 F.3d at 332. As we were careful to point out, the INS agents did not act on their own discretion but instead sought approval from numerous Department of Justice officials, "all of whom answer directly to the Attorney General herself." Id. at 334.

The logic of the conclusion that federal law enforcement officers must obtain Cabinet-level authorization for making extraterritorial arrests finds persuasive expression, in fact, in a legal opinion issued by the Department of Justice's own Office of Legal Counsel ("OLC"), legal advisor to the Attorney General and to the Executive branch generally. In 1989, Assistant Attorney General William P. Barr addressed the very questions we confront here, but in the context of the FBI's authority "to investigate and arrest individuals for violating United States law, even if the FBI's actions contravene customary international law." 13 U.S. Op. Off. Legal Counsel 163 (1989) ("Barr Opinion"). The Barr Opinion contended — contrary to a previous 1980 OLC Opinion — that the FBI had such authority, either statutorily or at least through the Attorney General. Whatever weight the Barr Opinion merits, it did not endorse the FBI's ability to act on its own authority, but rather cautioned that the FBI may violate international law only "at the direction of the President or the Attorney General." Id. at 183. Indeed, the Opinion was even more explicit, and cautionary, in its advice regarding the extent and exercise of authority "to override customary international law" in, for example, forcibly abducting a suspect from another country without that country's consent. Id. at 180-81. The Barr Opinion advised in a prescient passage that:

[W]e believe that the Attorney General has the power to authorize departures from customary or other international law in the course of law enforcement activities and that the President need not personally approve such actions. We would not recommend, however, that the Attorney General delegate the authority to more subordinate officials. Even if he is viewed as exercising statutory authority . . . we think that as a prudential matter the Attorney General should, in this case, exercise it personally. Decisions such as Garcia-Mir [v. Meese, 788 F.2d 1446 (11th Cir. 1986)] rely on the theory that the Executive has the constitutional authority to make political decisions affecting our international relations. To the extent that such decisions are made by officials below cabinet rank, however, the factual basis for this theory may be weaker.

Specifically, we recommend that any overseas law enforcement activity that presents a significant possibility of departing from customary or other international law be approved directly by the President or the Attorney General.

Id. at 180 (emphasis added).

Judge O'Scannlain's dissent therefore misses the mark by repeatedly stating that Congress has delegated to the Attorney General the authority to determine whether to enforce our laws extraterritorially. Even if such a delegation has in fact occurred — which has not been demonstrated here — it would be of little import in this case, because there is no evidence that the Attorney General played a role in the decision to abduct Alvarez.

Given the absence of specific approval by the Attorney General or any other Cabinet-level official, Judge O'Scannlain argues that the DEA enjoys the more general statutory authority to make extraterritorial arrests on its own accord. In support of this contention, he relies on Chen, claiming that the DEA's authority to act extraterritorially is the same as that of the INS. The critical distinction, however, is that the Attorney General has explicitly delegated to the INS his broad powers to enforce the immigration laws.² See 8 C.F.R. § 2.1. In the case of the DEA, the Attorney General has made no such delegation.

2 Moreover, as I explained earlier, the INS agents who planned the operation in Chen sought high-level approval from the Department of Justice despite this general delegation of the Attorney General's authority.

Judge O'Scannlain's dissent urges that this lack of an explicit delegation does not matter because Congress has authorized the extraterritorial application of the criminal statutes for which Alvarez was charged, and this authorization "would seemingly sanction" the extraterritorial enforcement of those statutes. The extraterritorial application and the extraterritorial enforcement of criminal statutes are far from synonymous concepts, however. That Congress may have intended a criminal statute to reach conduct that occurs beyond our borders, and that United States courts would have jurisdiction over such crimes, does not mean that Congress also intended to give law enforcement officers unlimited authority to enforce the statute by entering a foreign nation, uninvited, to abduct a foreign national, in violation of international law. Indeed, that is why we enter into extradition treaties. It is therefore not enough to say, as Judge O'Scannlain contends, that "Congress must have intended to have the laws enforced [extraterritorially] by some member of the Executive branch," for even if Congress did so intend, I cannot conclude that Congress silently designated the DEA officials, rather than the Attorney General, as the Executive branch officials to whom it was entrusting the decision to engage in extraterritorial law enforcement.

In the wake of the brutal murder of DEA Agent Camarena, the Drug Enforcement Administration understandably wanted to capture and punish those who were responsible for the death of one of its own. But in the absence of congressional delegation of the authority to override another nation's territorial sovereignty — an absence that the majority opinion has amply demonstrated — the decision to sneak into a friendly nation and abduct one of it citizens, in violation of international law, was not for the DEA to make. That decision belonged to the Attorney General and other members of the Cabinet, if not to the President himself. Because the highest levels of the Executive branch played no role in planning or authorizing Alvarez's abduction, and because Congress has not granted the DEA the more general authority to conduct extraterritorial law enforcement activities, I agree that the arrest and detention of Alvarez were arbitrary because they were not "pursuant to law." Alvarez therefore has established a violation of the law of nations that is actionable under the ATCA.

O'SCANNLAIN, Circuit Judge, with whom Circuit Judges RYMER, KLEINFELD, and TALLMAN join, dissenting:

We are now in the midst of a global war on terrorism, a mission that our political branches have deemed necessary to conduct throughout the world, sometimes with tepid or even non-existent cooperation from foreign nations. With this context in mind, our court today commands that a foreign-national criminal who was apprehended abroad pursuant to a legally valid indictment is entitled to sue our government for money damages. In so doing, and despite its protestations to the contrary, the majority has left the door open for the objects of our international war on terrorism to do the same.¹

1 Perhaps cognizant that its analysis cannot bear its own weight if applied more broadly, the majority recites that we need not worry because its holding "is a limited one." Supra at 608. Count me, however, among those unassuaged by the majority's assurances. I believe that impermissibly encroaching upon the duties rightfully reserved to the political branches is of serious consequence, and unfortunately such encroachment establishes a very troubling precedent which we will regret. Indeed, the majority's attempt to distinguish the circumstances of this case from other overseas operations conducted by our nation's military and law enforcement personnel may not prove to be so facile. One of the many vexing questions implicated by its opinion, but left unanswered by the majority, is what are we to make of sub-agencies within the Department of Homeland Security, as well as the Federal Bureau of Investigation ("FBI"), the Drug Enforcement Administration ("DEA"), the Bureau of Alcohol, Tobacco, and Firearms ("ATF"), and other law enforcement agents who aid and assist in the war against terrorism and efforts to protect homeland security by capturing known terrorists and criminals in foreign locales across the globe? Unless the majority believes that every use of transborder arrest by the Executive branch falls within "its power to detain under the war powers of Article II," supra at 609 (quoting Hamdi v. Rumsfeld, 316 F.3d 450, 473 (4th Cir. 2003)) — which is obviously not the case-no rational observer can honestly say that our court's holding today "is a limited one."

What makes this astounding pronouncement even more perverse is that our court divines the entitlement to recovery from the Alien Tort Claims Act ("ATCA"), 28 U.S.C. § 1350, a statute first enacted over 200 years ago by members of the First Congress, many of whom were Framers of our nation's Constitution. With utmost respect to the majority, there is simply no basis in our nation's law for this bewildering result, and the implications for our national security are so ominous that I must dissent.

I

Notwithstanding the majority's lengthy disquisitions concerning various theories and sources of international law, the central issue in this case is very simple: Do American law enforcement agents violate well-established principles of American jurisprudence when they apprehend a duly-indicted suspect outside the confines of our nation's borders?² The answer is clearly no; the United States has neither now nor ever agreed to an asserted international law principle prohibiting the practice of transborder abduction.³

2 As an initial matter, I am sympathetic with many of the separation of powers concerns expressed in Judge Gould's separate dissent. Indeed, I share a similar apprehension that the majority's approach could have dire consequences if applied to our nation's current military and law enforcement operations overseas.
However, interestingly enough, the government, neither in its brief on cross-appeal nor its amicus brief, argued for the applicability of the political question doctrine. In any event, under our precedent, I believe that Alvarez, while not entitled to relief, has stated a justiciable claim under the ATCA.

3 I agree with the majority that Alvarez lacks standing to obtain redress under the ATCA for Sosa's and the DEA agents' alleged infraction against Mexican sovereignty; state-on-state injuries like the one Alvarez alleges here are singularly inappropriate for assertion of third-party rights by foreign citizens.
Moreover, I agree with the majority that Alvarez's claim for transborder abduction must fail. However, because the majority reaches this result in a rather circuitous manner, I write separately on this issue to underscore that the United States has neither acquiesced in, nor considers itself bound by, any supposed norm against transborder arrest.

The majority, perhaps overlooking the grandeur of the forest while gazing with much admiration at the trees, meanders through various sources which suggest how pleasant it would be if transborder abduction were actionable. However, the majority's searching inquiry into the scope of international law is simply unnecessary. The ATCA is a congressionally enacted statute; accordingly, international law in this context must first and foremost comport with American case law and congressional intent, rather than be defined by the amorphous expressions of other countries or international experts. In other words, no claim can be actionable under the ATCA based on a norm to which the United States itself does not subscribe.

I do not suggest that the majority's inquiry into the status of transborder arrest in the broader international community — which Congress, by enacting the ATCA, has directed us to perform in appropriate cases — is one beyond the federal courts' ability to undertake. Indeed, some areas of substantial international unanimity are easily recognized. See, e.g., Trajano v. Marcos (In re Estate of Ferdinand E. Marcos Human Rights Litig.) ("Marcos I"), 978 F.2d 493, 500 (9th Cir. 1992). Nevertheless, I believe that in many cases, as in this one, it will be far easier to determine whether the United States subscribes to a given norm than whether other countries do, and accordingly the former inquiry should appropriately precede the latter.

II

I respectfully suggest that the majority has imprudently ignored the relevant underpinnings of the ATCA. As demonstrated below, a proper historical understanding of the ATCA compels the conclusion that no claim can prevail where the United States, through its political branches, does not acquiesce in an international norm.

A

First enacted as part of the Judiciary Act of 1789, the ATCA still reads today almost exactly as the First Congress drafted it; the version currently enshrined in Title 28 provides: "The district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States." 28 U.S.C. § 1350 (1994); see Judiciary Act of Sept. 24, 1789, ch. 20, § 9(b), 1 Stat. 73, 77.

The ATCA was, from the beginning, a curious provision. As one eminent scholar of both federal jurisdiction and American legal history notes, the ATCA was one of only two provisions of the Judiciary Act that "arguably g[a]ve federal courts jurisdiction over judicial matters outside the enumeration of Article III." David P. Currie, The Constitution in Congress: The Federalist Period 1789-1801, at 51-52 (1997).⁴ Perhaps because of the singular nature of its jurisdictional grant, the ATCA was infrequently used for almost two hundred years, until fairly recently when courts have eagerly exploited the opportunity to revivify it.

4 The other was the Act's apparent provision for general alienage diversity jurisdiction, rather than jurisdiction only over controversies between aliens and U.S. states or citizens, as specified in Article III, Section 2. Currie, supra, at 51. The Supreme Court subsequently construed the statute's reference to suits "where an alien is a party" to comprehend only suits "between citizens and foreigners," to conform to the Article III grant. Mossman v. Higginson, 4 U.S. (4 Dall.) 12, 14, 1 L.Ed. 720 (1800) (emphasis omitted).

In the course of this resurgence of a statutory provision that lay largely dormant since our nation's founding, our court has determined that certain international law principles may be incorporated into federal common law, and thereby into the ATCA as well. See Hilao v. Estate of Marcos (In re Estate of Ferdinand Marcos, Human Rights Litig.) ("Marcos II"), 25 F.3d 1467, 1475 (9th Cir. 1994). The Marcos II court set out the standard for evaluating whether an ATCA plaintiff states a claim: "Actionable violations of international law must be of a norm that is specific, universal, and obligatory." Id; accord, e.g., Papa v. United States, 281 F.3d 1004, 1013 (9th Cir. 2002); Martinez v. City of Los Angeles, 141 F.3d 1373, 1383-84 (9th Cir. 1998).

B

The requirement of "universality" constitutes an insurmountable bar to recovery for transborder arrest.⁵ I focus in particular on the corollary of this requirement: a norm of international law not recognized by the United States cannot be deemed a universal one, actionable in this nation's courts.

5 Of the elements required for an actionable norm under the ATCA, "specificity" is, appropriately, the clearest of the three. International law from the time of the ATCA's enactment has been somewhat inchoate, and as the number of international agreements, conventions, and organizations has grown, discerning the substance of the law of nations has required rather more than reading the works of Pufendorf, Burlmaqui, and Vattel with which members of the First Congress were presumably familiar. Moreover, the international community whose customs and practices define the law of nations has become larger and more diverse. It is not surprising, therefore, that frequently the propositions capable of attracting the broadest support are also the most diffuse (and thus the least likely to offend). Yet much diplomatic gloss, though possessing great virtue for its significance to the development of the law of nations in the broadest sense, provides no suitable basis for tort litigation.
A "specific" norm, therefore, is one sufficiently "'definable,'" Marcos II, 25 F.3d at 1475 (quoting with approval Forti v. Suarez-Mason, 672 F.Supp. 1531, 1539-40 (N.D.Cal. 1987)), such that its violation can be objectively ascertained. To be sure, the nations of the world need not have commonly agreed upon an exhaustive catalogue of every variation, but the norm itself must have become "clear and unambiguous." Tel-Oren v. Libyan Arab Republic, 726 F.2d 774, 819-20 (D.C. Cir. 1984) (Bork, J., concurring) (quoting Filartiga v. Pena-Irala, 630 F.2d 876, 884 (2d Cir. 1980)) (internal quotation marks omitted).

We have previously noted the importance of determining whether a norm of international law is recognized by the United States. See Martinez, 141 F.3d at 1383 ("To determine whether this tort satisfies the requirement for a tort claim under the Alien Tort Act, we must decide '[1] whether there is an applicable norm of international law [proscribing such a tort] . . . recognized by the United States . . . and [2] whether [that tort] was violated in [this] particular case.'" (quoting Marcos I, 978 F.2d at 502 (alterations in original) (emphasis added))). Marcos I did not state this requirement explicitly, but the exposition of the constitutional basis for the ATCA, see supra at 612, makes clear that the Martinez court correctly recognized that ATCA jurisdiction subsumes it.

Federal common law is a means of preserving a uniform national construction of rights and obligations within a given area of the law even in the absence of a detailed statutory scheme. See, e.g., Tex. Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 640-41, 101 S.Ct. 2061, 68 L.Ed.2d 500 (1981). This consideration carries particular force in the foreign policy context in which the ATCA lies; it was passed, let us remember, in 1789, only months after the First Congress convened. The Framers, and presumably those who went on to serve the new government, were acutely conscious of the need for the national government's interpretation of the law of nations to be controlling. See, e.g., The Federalist No. 3, at 43 (John Jay), No. 80, at 476-77, 478 (Alexander Hamilton) (Clinton Rossiter ed., 1961). Yet one equally basic characteristic of federal common law is that Congress may supplant it as the rule of decision, because the power to legislate rests most properly with the elected representatives who possess both the greater competence and the greater authority, conferred by the people, to wield it. And the same is no less true with regard to the law of nations as federal common law; indeed, foreign policymaking is essentially confided not merely to the national government writ large, but to its political branches in particular.⁶ E.g., Chi. S. Air Lines, Inc. v. Waterman S.S. Corp., 333 U.S. 103, 111, 68 S.Ct. 431, 92 L.Ed. 568 (1948); Oetjen v. Cent. Leather Co., 246 U.S. 297, 302, 38 S.Ct. 309, 62 L.Ed. 726 (1918); see Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 423, 84 S.Ct. 923, 11 L.Ed.2d 804 (1964).

6 Of course, accepting this principle still leaves open the question of how that responsibility should be allocated between the political branches. For present purposes, it is sufficient to say that whatever degree of responsibility the President enjoys vis-à-vis Congress, the political branches collectively enjoy primacy over the judiciary in the management of our foreign relations.

The Framers and the First Congress viewed the United States's substantial adoption of the law of nations as furthering their intention that the new nation take its place among the civilized nations of the world. E.g., Chisholm v. Georgia, 2 U.S. (2 Dall.) 419, 474, 1 L.Ed. 440 (1793) (opinion of Jay, C.J.). Yet they clearly did not mean for the law of nations to act as an irrevocably binding constraint on the law and policy-making authority of the national government. In his last contribution as Publius, John Jay famously recognized the binding nature of treaties, and a number of the Framers shared his view that treaties created a binding obligation on the contracting parties under the law of nations. The Federalist No. 64, at 394 (John Jay); see, e.g., Note, Restructuring the Modern Treaty Power, 114 Harv. L. Rev. 2478, 2484-90 (2001) (discussing the Framers' views). But the law of nations imposed constraints "in point of moral obligation," not restrictions on national policymakers' power to breach. Ware v. Hylton, 3 U.S. (3 Dall.) 199, 272, 1 L.Ed. 568 (1796) (opinion of Iredell, J.). The Framers intended that the United States would have the power, if not necessarily the right under the law of nations, to violate or even to repudiate aspects of the law of nations, provided it were willing to face the consequences of its breach, possibly including war.⁷ And it was for this reason that the power to violate or to repudiate, like the power over foreign affairs generally, was confided to the national government. See The Federalist No. 80, supra, at 476 ("[T]he peace of the WHOLE ought not to be left at the disposal of a PART. The Union will undoubtedly be answerable to foreign powers for the conduct of its members. And the responsibility for an injury ought ever to be accompanied with the faculty of preventing it."). The federal common law's incorporation of the law of nations, in short, is not beyond the political branches' power to alter. And this fact is entirely consonant with the principle, expressed in our cases as elsewhere, that "[c]ustomary international law, like international law defined by treaties and other international agreements, rests on the consent of states." Siderman de Blake v. Republic of Argentina, 965 F.2d 699, 715 (9th Cir. 1992); see id. ("A state that persistently objects to a norm of customary international law that other states accept is not bound by that norm. . . ." (citing Restatement (Third) of the Foreign Relations Law of the United States [hereinafter Restatement] § 102 cmt. d)).

7 See, e.g., Louis Henkin, International Law as Law in the United States, 82 Mich. L. Rev. 1555, 1568 (1984) ("[E]very State has the power — I do not say the legal right — to denounce or breach its treaties, or to violate obligations of customary international law." "[I]t is inconceivable that the Constitution intended to make it impossible or impermissible — unconstitutional — for the United States to violate a treaty or other international obligation."). One of the handful of cases sustaining ATCA jurisdiction, Bolchos v. Darrel, 3 F.Cas. 810 (D.S.C. 1795) (No. 1,607), likewise recognizes the power to depart from the law of nations. See id. at 811 ("It is certain that the law of nations would adjudge neutral property, thus circumstanced, to be restored to its neutral owner; but the . . . treaty with France alters that law. . . ."). Although the law of nations did not favor the libelant, the Bolchos court found in his favor nonetheless because jurisdiction was proper under the ATCA's treaty provision and because the treaty in question conferred an enforceable right to sue. See id. Despite Bolchos's distasteful subject matter — the action was for recovery of a cargo of slaves — the opinion offers some insight into contemporaneous understanding of the law of nations as enforced by the ATCA.

Therefore, where the political branches have exercised their power to diverge from the course that others see the law of nations as setting, ATCA liability cannot lie. The ATCA's conformity with Article III rests on the incorporation of the law of nations as federal common law — particularly in a case like this one, where neither alienage, nor admiralty, nor any of the other headings of Article III provides a basis for federal jurisdiction. It is for this reason that an ATCA plaintiff relying on the law of nations (as opposed to a treaty) must allege a tort that violates some norm of international law recognized by the United States.

At least one of the few Supreme Court opinions to consider the ATCA directly appears to have recognized as much. In O'Reilly De Camara v. Brooke, 209 U.S. 45, 28 S.Ct. 439, 52 L.Ed. 676 (1908), Justice Holmes wrote for a unanimous Court in affirming the dismissal of an ATCA complaint that alleged the tortious destruction of a hereditary title during the Spanish-American War. Id. at 48-49, 28 S.Ct. 439. Although the Court did not directly decide whether the plaintiff had alleged a tort cognizable under the ATCA, see id. at 52-53, 28 S.Ct. 439, it did make the following comment on that question: "[W]e think it plain that where, as here, the jurisdiction of the case depends upon the establishment of a 'tort only in violation of the law of nations, or of a treaty of the United States,' it is impossible for the courts to declare an act a tort of that kind when the Executive, Congress, and the treaty-making power all have adopted the act." Id. at 52, 28 S.Ct. 439.

Such analysis fits with our case law's incorporation of the requirement that an actionable norm of international law be "universal." The case at hand does not require us to delineate the bare minimum level of acceptance that would constitute "universality"; instead, it invokes the simple proposition, stated explicitly in Martinez and implicitly in other cases, that in determining whether a norm is "universal," the United States is to be counted as a part of the universe. A norm to which the political branches of our government have refused to assent is not a universal norm. It is not the judiciary's place to enforce such a norm contrary to their will.⁸

8 The final requirement under our law is that an actionable norm be "obligatory." Our cases have used this term to mean "binding" rather than merely "hortatory." Marcos II, 25 F.3d at 1475 (quoting with approval Forti, 672 F.Supp. at 1539-40). Binding norms "confer fundamental rights upon all people vis-a-vis their own governments." Id. at 1475-76 (quoting Filartiga, 630 F.2d at 885) (alteration in original) (internal quotation marks omitted). Sosa and the DEA agents argue that the class of obligatory norms is further restricted to those that are obligatory in the literal sense, i.e., those that nation-states must obey. In the parlance of international law, such a norm falls under the heading of jus cogens.
This argument is not without support in the case law. Courts applying Marcos II's tripartite standard have debated whether the only "obligatory" norms whose violation is actionable under the ATCA are those that have attained jus cogens status, and indeed, some have adopted the position that the defendants urge. Compare, e.g., Nat'l Coalition Gov't of the Union of Burma v. Unocal, Inc., 176 F.R.D. 329, 345 (C.D.Cal. 1997) ("[A] foreign sovereign's expropriation of its national's property does not constitute a jus cogens violation of the law of nations and, therefore, is not cognizable under § 1350."), Xuncax v. Gramajo, 886 F.Supp. 162, 184 (D.Mass. 1995) ("These qualifications essentially require that . . . the prohibition against [the act in question be] non-derogable and therefore binding at all times upon all actors."), and Beanal v. Freeport-McMoRan Copper Gold, Inc., 969 F.Supp. 362, 370 (E.D.La. 1997) (citing Xuncax), aff'd, 197 F.3d 161 (5th Cir. 1999), with Doe v. Unocal Corp., 110 F.Supp.2d 1294, 1304 (C.D.Cal. 2000) ("While the Ninth Circuit has not expressly held that only jus cogen [ sic] norms are actionable, the Circuit's holding in Estate II that actionable violations are only those that are specific, universal, and obligatory is consistent with this interpretation."), and In re World War II Era Japanese Forced Labor Litig., 164 F. Supp.2d 1160, 1179 (N.D.Cal. 2001) ("It remains unclear, however, whether all jus cogens norms meet the 'specific, universal and obligatory standard' required to be actionable under section 1350."), with Mehinovic v. Vuckovic, 198 F.Supp.2d 1322, 1344 (N.D.Ga. 2002) ("A jus cogens violation satisfies, but is not required, to meet [the 'specific, universal, and obligatory'] standard.").
However, we need not decide whether the term "obligatory" necessarily means that a jus cogens norm is required. Our prior cases have left this question open. And despite the majority's assertion to the contrary, this case does not require us to pronounce definitively on the relevance of the distinction between jus cogens and mere customary international law. If Alvarez in fact alleges violations of a jus cogens norm, as did the plaintiffs in the Marcos litigation, or if he alleges violations of a derogable norm to which the United States does not subscribe, we may bypass the significance of jus cogens status. The distinction would be significant only if Alvarez alleges violations of a norm that is recognized by the United States — and therefore incorporated into the federal common law — but lacks the universal acceptance within the international community that is the sine qua non of a jus cogens norm.

C

The majority fails to take into account these fundamental principles, which invoke the historical understanding in which the ATCA was passed and the proper role of the political branches in determining our nation's actions as they relate to national security. As a result, the majority's analysis of the status of transborder arrest as an instrument of law enforcement — which has great bearing on its treatment of Alvarez's alternative claim for relief for arbitrary arrest — is seriously flawed because it ignores the baseline proposition that it is easier for this court to determine whether the United States agrees with a norm than to determine whether a preponderance of the world's other nations does.

Let us recall the particular circumstances of the case at hand. Alvarez was charged in 1990 with, among other offenses, the kidnaping⁹ and felony murder of a federal agent in violation of 18 U.S.C. §§ 1114(1) and 1201(a)(5). See, e.g., Alvarez-Machain II, 504 U.S. at 657 n. 1, 112 S.Ct. 2188. The kidnaping statute appears to contemplate the exercise of federal criminal jurisdiction over certain defendants who are neither Americans nor found in the United States. See 18 U.S.C. § 1201(e) (2000) (authorizing the exercise of jurisdiction over defendants if they are American nationals, if they are found in the United States, or "if the victim is a representative, officer, employee, or agent of the United States"). Notwithstanding the majority's assertions to the contrary, the general statutes governing the operation of the Drug Enforcement Administration confer on DEA agents the authority to "make arrests without warrant . . . for any felony, cognizable under the laws of the United States, if [an agent] has probable cause to believe that the person to be arrested has committed or is committing a felony." 21 U.S.C. § 878(a)(3) (2000). DEA agents may also "perform such other law enforcement duties as the Attorney General may designate." Id. § 878(a)(5).

9 A 1994 amendment added a second "p" to the term, which now lists the crime as "kidnapping."

This statutory framework confers on the DEA agents the same degree of authority to act extraterritorially that we have previously held INS agents to possess. See United States v. Chen, 2 F.3d 330 (9th Cir. 1993). In Chen, we considered whether the INS was authorized to conduct criminal law enforcement activity outside the United States. We held that "Congress need not confer such authority explicitly and directly on the INS agents themselves." Id. at 333. We inferred from Congress's broad grant of authority to the Attorney General to enforce the immigration laws,¹⁰ and from the extraterritorial applicability of those laws, that the enforcement power extended where the laws themselves extended. See id. at 333-34.

10 She had in turn delegated that authority to the Commissioner of Immigration and Naturalization, who in turn delegated it to rank-and-file INS agents. Chen, 2 F.3d at 334 (citing 8 C.F.R. § 2.1 (1991)).

Faced with our holding in Chen, Alvarez argues, and the majority erroneously agrees, that the criminal context presented in this case is distinguishable from the unique context of border security at issue in Chen. Yet the statutes at issue here bear just as directly on national security, particularly insofar as they relate to and promote the federal government's ability to enforce the drug laws, and to protect the agents who carry out that enforcement, no matter on which side of the border they may be threatened. And so our cases have recognized. "Our circuit has repeatedly approved extraterritorial application of statutes that prohibit the importation and distribution of controlled substances in the United States because these activities implicate national security interests and create a detrimental effect in the United States." United States v. Vasquez-Velasco, 15 F.3d 833, 841 (9th Cir. 1994). As we stated in another prosecution arising from the Camarena abduction and murder: "We have no doubt that whether the kidnapping and murder of [DEA] agents constitutes an offense against the United States is not dependent upon the locus of the act. We think it clear that Congress intended to apply statutes proscribing the kidnapping and murder of DEA agents extraterritorially." United States v. Felix-Gutierrez, 940 F.2d 1200, 1204 (9th Cir. 1991).

Applying the reasoning of Chen to the statutes that protect American drug enforcement personnel leads to the inescapable conclusion that Congress has authorized federal agents enforcing those statutes to make warrantless arrests beyond our borders — a conclusion at odds with the argument that the United States respects a norm prohibiting transborder arrests.

Moreover, such a norm, which would render a transborder arrest violative of the law of nations absent the host country's consent, does not seem tenable either as a matter of statutory construction or as a reflection of Congress' likely goal. I can conceive of a number of situations in which the nature of the host country's government, or even the utter nonexistence of a functioning government, precludes obtaining the formal sanction of the local judiciary or of the host country. Indeed, in the months since September 11, 2001, the United States government has retrieved a number of individuals from lawless locales of this sort. See, e.g., Al Odah v. United States, 321 F.3d 1134 (D.C. Cir. 2003). Therefore, the statutory authorization to make arrests overseas for violations of extraterritorially applicable law runs contrary to an alleged prohibition on transborder arrests.

It is true that in a number of other cases dealing with transborder arrests, the nation in which the arrest occurred did not object, even if it did not cooperate. While Mexico, by contrast, clearly and consistently protested Alvarez's seizure, it does not follow that the United States was obliged to comply with the processes of Mexico's judicial system.

One highly visible abduction, that like the case at hand also involved the extraterritorial enforcement of our nation's drug laws, refutes the notion that the United States has somehow divested itself of the authority to engage in transborder arrest without the consent of the host country. United States v. Noriega, 117 F.3d 1206 (11th Cir. 1997), of course, dealt with the arrest of Panama's former strongman on drug-related charges. Id. at 1209-10.¹¹ Noriega was functioning "as the de facto, if not the de jure, leader of Panama" when the American military incursion to seize him occurred. Id. at 1211. Surely any protest from a Panamanian government controlled by Noriega himself could be disregarded, could it not, particularly if, as a matter of law, Noriega had forfeited his head-of-state immunity? See id. at 1212. Moreover — and herein lies the rub — the United States had consistently refused to acknowledge the legitimacy of Noriega's rule from its inception, which was after his indictment but well before the effort to retrieve him was ordered. Id. at 1209-10.

11 For purposes of further illustration, I refer to the recent case of Kasi v. Angelone, 300 F.3d 487 (4th Cir. 2002). Mir Aimal Kasi on the morning of January 25, 1993, stopped his automobile behind a line of automobiles outside of CIA headquarters in Fairfax County, Virginia, emerged from his vehicle, and opened fire on the other drivers with an AK-47 assault rifle. Two CIA employees were killed, and three other employees were wounded. Kasi fled to his home country of Pakistan the day after the shootings in order to avoid arrest. On February 16, 1993, Kasi was indicted for the various crimes that he had committed. For the next four and one-half years, Kasi remained at-large, traveling in Afghanistan and returning to Pakistan only for brief visits. Then, on June 15, 1997, FBI agents located and abducted Kasi from a hotel in Pakistan. Two days later, Kasi was transported back to Fairfax Country, Virginia and handed over to state authorities for prosecution. Kasi was eventually convicted and sentenced to death for his crimes. In response, therefore, to the majority's assertion that not "every executive branch decision to breach an international norm translates into a more global repudiation of that norm," supra at 620 n. 15, Kasi, Noriega, and Chen, as well as the recent capture of terrorists all across the globe, see, e.g., Al Odah v. United States, 321 F.3d 1134 (D.C. Cir. 2003), are merely a few examples that demonstrate the consistent refusal of the political branches of this nation's government to adhere to a prohibition against transborder arrests. Consequently, the ATCA should provide no monetary relief to those criminals and terrorists captured pursuant to this valid tool of national security.

Therefore, irrespective of what various international law scholars and others may deem as advisable policy, these ruminations are of little consequence under the ATCA when the political branches of the United States have firmly decided on a course of action. Examining the relevant statutes, the actions of the political branches in other circumstances, and as discussed by the majority, the manner in which the President and the Senate have exercised the treaty power in this area, see supra at 618-619, leads to only one conclusion: The United States does not, as a matter of law, consider itself forbidden by the law of nations to engage in extraterritorial arrest, but reserves the right to use this practice when necessary to enforce its criminal laws.¹²

12 Because the prohibition on transborder arrest is not accepted by the United States and thus is not actionable under the ATCA, it is unnecessary to consider whether, as Sosa and the DEA agents urge, it is not actionable for the additional reason that it does not rise to the level of jus cogens.
However, it should be noted that no court, convention, declaration, or authority such as the Restatement has identified transborder abduction as a jus cogens norm. Rather Ker v. Illinois, 119 U.S. 436, 7 S.Ct. 225, 30 L.Ed. 421 (1886), Alvarez-Machain II, 504 U.S. 655, 112 S.Ct. 2188, 119 L.Ed.2d 441 (1992), and United States v. Verdugo-Urquidez, 494 U.S. 259, 110 S.Ct. 1056, 108 L.Ed.2d 222 (1990) indicate the opposite. See United States v. Matta-Ballesteros, 71 F.3d 754, 763 n. 5 (9th Cir. 1995) (kidnaping does not violate recognized constitutional or statutory provisions in light of Alvarez-Machain II, and also does not qualify as a jus cogens norm such that its commission would be justiciable in our courts even absent a domestic law). Indeed, on remand in this case, we made it clear that there was no due process violation — put differently, that kidnaping isn't shocking — because it cannot be so fundamental as to constitute a jus cogens norm. Alvarez-Machain III, 107 F.3d 696, 702 (9th Cir. 1997).
The Restatement, for example, has identified only torture, genocide, slavery, murder, prolonged arbitrary detention, and systematic racial discrimination. Neither do the conventions and declarations upon which Alvarez relies establish universal acceptance. The International Covenant on Civil and Political Rights was signed and ratified in 1992 but with the understanding by the Senate and Executive Branch that Articles 9 and 10 are not self-executing and may not be relied on by individuals; the American Convention on Human Rights has not been ratified; the Universal Declaration of Human Rights and the American Declaration of the Rights and Duties of Man are not binding legal obligations. None prohibits forcible abduction. Nor does it suffice to rely on general principles such as "rights to freedom of movement, to remain in one's country, and security in one's person."

III

If the majority had merely denied Alvarez's claims based on its exposition of international mores, I would be troubled by its failure to engage in a review of both the history behind the ATCA and the manner in which the political branches have exercised the option of transborder arrest in varied circumstances. But at least I would feel secure that we as judges had not improperly encroached upon the duties reserved for the political branches in formulating our nation's foreign policy. Most regrettably, by providing relief to Alvarez on his claim of prolonged arbitrary arrest, our court has in effect restricted the authority of our political branches, and it has done so in a way that finds no basis in our law.

A

The parties do not dispute that the prohibition on prolonged arbitrary detention is actionable under the ATCA. See Martinez, 141 F.3d at 1383-84. Nevertheless, the question remains whether the prohibition was violated in this case.

B

Notwithstanding its express recognition that the criminal statutes covering kidnaping, felony murder, and the other crimes involved in this case extend to conduct outside of the borders of the United States, the majority deems the government's action as "arbitrary."

The majority, although not expressly stating it as such, seems most troubled by the lack of Mexican authority for Alvarez's arrest. Indeed, imagine that the DEA had communicated with the Mexican government prior to seizing Alvarez and that such dialogue led to Mexican authorities assisting in the arrest, or acknowledging consent to the DEA's actions in some other manner. Under the majority's proffered approach the United States would be forced to compensate an alleged foreign-national criminal for "arbitrary arrest" within the meaning of the law of nations merely because the "wrong" Executive agency spearheaded the operation.

If Mexico had indeed sanctioned its actions, our court would not be subjecting the DEA to liability for successfully negotiating via diplomatic means the capture of a wanted criminal. Or would it? This simple hypothetical, however, underscores the fallacy of the majority's approach. Whether the United States procured an arrest warrant through the Mexican judiciary should not affect our analysis, because the availability of local process is extremely sensitive, bound up with important foreign policy considerations that are confided to the political branches in general and to the Executive in particular. Indeed, the majority elsewhere recognizes that "an individual's right to be free from transborder abductions has not reached a status of international accord sufficiently to render it 'obligatory' or 'universal,' [and therefore] cannot qualify as an actionable norm under the ATCA." Supra at 620.

Nevertheless, seemingly under the majority's approach, such extraterritorial arrest authority may still be subject to a requirement that any agents exercising it on foreign soil — as opposed to the high seas, as was the case in Chen, 2 F.3d at 332 — obtain the consent or assistance of the host country. The majority, in a rare moment of restraint, does recognize the "powers of the political branches to override the principles of sovereignty in some circumstances, should the need arise." Supra at 629 (emphasis added). However, the "need" to engage in transborder arrest without the prior consent of a foreign nation should appropriately be left to the discretion of the political branches.

Indeed, the federal courts are not charged with determining the legitimacy of another nation's government. Yet this is essentially what the majority would have us do. Under its approach, the United States would have departed from the ostensibly black-and-white approach that sanctions transborder arrests when employed by our nation's political branches. In its place, a decision to make a transborder arrest would only be permissible when the host country's system of government absolutely requires it, as determined by this country's courts through the medium of ATCA litigation. As judges, we would have to determine whether a nation's courts are open and functioning; whether it has a legitimate government that can be consulted for permission to seize a suspect; if there are multiple contenders, see, e.g., Noriega, 117 F.3d at 1209-10, to which one such a request must be addressed; and so on. Courts are quite unsuited to undertake such analyses, and, indeed, to do so would bring us perilously close to trenching on the power of diplomatic recognition that Article II, Section 3 places at the core of the Executive's foreign affairs authority. See, e.g., Guaranty Trust Co. v. United States, 304 U.S. 126, 137-38, 58 S.Ct. 785, 82 L.Ed. 1224 (1938); Oetjen, 246 U.S. at 302-03, 38 S.Ct. 309.

I am simply not prepared to declare that Congress intended that any alien charged with a crime, under extraterritorially applicable U.S. criminal law, could remain in a country that refuses extradition. Congress has authorized the arrest, without warrant, of aliens for whom there is probable cause to suspect violation of an extraterritorially applicable statute. In so doing, Congress has left to the Executive, which already possesses the general responsibility for deciding both when and whether to arrest and to prosecute and how best to conduct the nation's foreign relations, the burden of determining when the national interest requires bypassing diplomatic channels to secure such arrest. As the Supreme Court has held in another context, "Situations threatening to important American interests may arise half-way around the globe, situations which in the view of the political branches of our Government require an American response with armed force. If there are to be restrictions on searches and seizures which occur incident to such American action, they must be imposed by the political branches through diplomatic understanding, treaty, or legislation." United States v. Verdugo-Urquidez, 494 U.S. 259, 275, 110 S.Ct. 1056, 108 L.Ed.2d 222 (1990).

C

Turning now to its proffered reason for granting Alvarez redress under the ATCA, the majority claims there was "no basis [under United States] law for the DEA's actions." Supra at 631.

1

The majority reaches this extraordinary result even though it concedes that the United States has reserved for itself the authority to arrest criminals and terrorists abroad as a valid law enforcement technique, and that Congress has explicitly extended the reach of criminal statutes for which Alvarez was charged to apply to conduct outside of the nation's borders. Even more astonishing is that the majority bases its holding on the premise that the DEA's actions were "arbitrary" within the meaning of the law of nations because they were beyond the scope of authority conferred by Congress.

In this case, Alvarez was arrested pursuant to an American warrant, issued following his indictment by a federal grand jury on felony charges. He was held overnight and then brought to the United States, where he was promptly placed in federal custody and was arraigned as soon as his medical condition permitted. The majority fails to explain adequately how an arrest supported by probable cause and ordered by a warrant, leading to a brief period of confinement before transfer to custody on American soil with all its attendant legal process, rises to the level of arbitrary detention merely because a parallel warrant was not obtained from the harboring state.

This view does not necessarily render such a seizure legal in every respect; we are limited here to the question whether the arrestee can recover in tort under the Alien Tort Claims Act, which presupposes a violation of the law of nations.¹³ Whatever false arrest claim Alvarez might have, he has not stated a violation of the law of nations to which we adhere.

13 It is important to note that Alvarez did bring a number of other claims which, if proven, might well have been cognizable — under the ATCA or otherwise. He alleged that he suffered cruel and degrading treatment, that he was subjected to assault and battery, that his captors had intentionally inflicted emotional distress. The district court took several of these claims to trial and resolved each of them in Sosa's favor, finding that Alvarez was not credible. The only ATCA claims that survived for appeal were those relating to the undisputed fact of Alvarez's seizure, not his allegations of abuse, torture, or mistreatment. But in the appeal before us, it is clear that the mere fact of a transborder arrest without the host country's consent is not actionable under the ATCA, absent a substantiated claim of mistreatment that independently violates aspects of the law of nations that the United States recognizes.

2

In any event, contrary to the majority's surmise, the DEA was well within its delegated powers when arresting Alvarez. The relevant statutory provisions confer on DEA agents the authority to "make arrests . . . for any felony, cognizable under the laws of United States" and the added authority to "perform such other law enforcement duties as the Attorney General may designate." 21 U.S.C. § 878(a) (emphasis added). Because it is undisputed that Congress has authorized the extraterritorial application of the criminal statutes for which Alvarez was charged, see supra at 624, this broad legislative delegation of enforcement powers to the DEA would seemingly sanction the extraterritorial arrests at issue in this case.

Nevertheless the majority would narrow this broad delegation of enforcement power and restrict the DEA's authority to engage in transborder arrests because it concludes that it would be "anomalous" for Congress to confer a similar degree of authority to "any State or local law enforcement officer designated by the Attorney General." Supra at 626-627. However, there is nothing "anomalous" about the legislative branch delegating to the Attorney General to determine in his best judgment whether non-federal law enforcement agents can aid in the application and enforcement of this nation's criminal laws extraterritorially. Instead, Congress engaged in such a broad delegation of law enforcement authority to the DEA and to the Attorney General in order to allow the Executive branch to have the widest array of enforcement options at its disposal.

In addition to the clear language of 21 U.S.C. § 878(a), this court's statutory interpretation is guided by the Supreme Court's recognition that "Congress simply cannot do its job absent an ability to delegate power under broad general directives." Mistretta v. United States, 488 U.S. 361, 372, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989). Furthermore, "[d]elegation of foreign affairs authority is given even broader deference than in the domestic area." Freedom to Travel Campaign v. Newcomb, 82 F.3d 1431, 1438 (9th Cir. 1996). "Congress — in giving the Executive authority over matters of foreign affairs — must of necessity paint with a brush broader than it customarily wields in domestic areas." Id. (quoting Zemel v. Rusk, 381 U.S. 1, 17, 85 S.Ct. 1271, 14 L.Ed.2d 179 (1965)).

Nevertheless, the majority claims there is an utter void of authority for the DEA's actions. The majority is undeterred by the fact that Congress has authorized the extraterritorial application of the criminal statutes involved in this case and has delegated broad powers of enforcement to the DEA "for any felony, cognizable under the laws of the United States." The majority's holding flies in the face of the clear statutory language enacted by Congress as well as the principle of statutory construction that delegations to the Executive branch are entitled to greater judicial deference in matters involving politically sensitive foreign affairs.

Furthermore the majority's approach leads unavoidably to the following question: if Congress through enactment of 21 U.S.C. § 878(a) has not in fact authorized the DEA and Attorney General to enforce extraterritorially the criminal laws for which Alvarez was charged, to whom exactly has Congress delegated this enforcement authority? By extending the reach of our criminal laws to apply to conduct outside of the nation's borders, Congress must have intended to have the laws enforced by some member of the Executive branch.¹⁴

14 In a separate concurrence, Judge Fisher seemingly concedes that Congress has authorized the Executive branch to engage in extraterritorial law enforcement activities, but nonetheless divines that such decisions have to be made by the "President, the Attorney General, the Secretary of State, perhaps the Secretary of Defense, the National Security Advisor." Supra at 642 (Fisher, J., concurring). However, in matters of foreign affairs, nation-states are often intentionally guarded as to precisely which government officials authorize and possess knowledge of covert operations. Furthermore, regardless of whether it is desirable public policy to require high-ranking officials to admit their participation in complex international operations in order to shield the United States and its "sub-Cabinet-level enforcement officials" from liability under the ATCA, there is simply no basis in the Constitution, existing statutes, or our case law for such a legal conclusion.
Judge Fisher tries to find support for his position from a 1989 advisory opinion issued by an attorney in the Department of Justice's Office of Legal Counsel. This advisory opinion expressly recognized the authority of the Attorney General to depart from the norms of international law in the course of law enforcement activities, even without the approval of the President. The opinion then "recommend[s], however, that the Attorney General [not] delegate the authority to more subordinate officials." Supra at 642 (Fisher, J., concurring).
As a matter of public policy, this advisory opinion may be of interest to show how the Executive branch might exercise the discretion conferred to it by Congress to engage in extraterritorial law enforcement activities. But such an advisory opinion, of course, neither constitutes binding law nor justifies the newly minted judicial constraints proposed by Judge Fisher to be imposed upon the Executive branch for conducting foreign affairs.

Under the majority's approach, this nation would be left to the whims of foreign countries in enforcing its laws because Congress, in delegating broad law enforcement powers to the Executive branch, did not redundantly recite that extraterritorial enforcement is to be included.¹⁵

15 While today's holding may seem innocuous with regard to Mexico, "an important ally and trading partner," the current war on terror in Afghanistan and in other countries, Kasi, and Noriega are but a few examples that illustrate that such an approach would unduly interfere with the Executive branch's ability to carry out its prescribed duties regarding law enforcement and national security.

3

Both the statutory structure and our own precedent indicate that the criminal provisions in question apply extraterritorially. Correspondingly, the statutes and precedent also indicate that Congress has authorized their extraterritorial enforcement. Under such circumstances, we are not free to conclude that the political branches have bound themselves — or, to be more precise, have bound the Executive — to the mast. And the prospect of international opprobrium is not sufficient for us as judges to impose a constraint the political branches have not.¹⁶

16 The majority also concludes, as to Alvarez's Federal Tort Claims Act ("FTCA") claims, that the DEA agents authorized a false arrest against Alvarez. That Congress intended several of the federal statutes that Alvarez was charged with violating to be both applicable and enforceable beyond the borders of the United States, effectively begins and ends this inquiry. The federal officers were authorized by statute to make warrantless arrests. And the provision in Rule 4(d)(2) of the Federal Rules of Criminal procedure that a "warrant may be executed . . . at any place within the jurisdiction of the United States" — a provision that, when adopted, was intended to broaden the territorial scope of a federal court's power to issue process, see Fed.R.Crim.P. 4 advisory committee's note — does not preclude federal authorities from obtaining an arrest outside the United States when Congress has authorized them to enforce an extraterritorially applicable statute beyond the national borders.
In light of the statutory authority under which the agents secured Alvarez's arrest in Mexico, the contention that their action was compensable in tort under the FTCA must fail. The district court did not err in granting summary judgment to the United States on the FTCA claims.

IV

Dr. Alvarez's capture and delivery to the United States may have offended the sensibilities of some members of our court. As a matter of public policy, such actions may even be worthy of the condemnation that certain pundits and foreign countries, as cited by the majority, have bestowed. But we are not asked in this case to condemn or to condone the federal government's actions; we are asked to compensate Dr. Alvarez in tort under the law of nations. The decision to exercise the option of transborder arrest as a tool of national security and federal law enforcement is for the political branches to make. They, unlike the courts, may be held accountable for any whirlwind that they, and the nation, may reap because of their actions. By its judicial overreaching, the majority has needlessly shackled the efforts of our political branches in dealing with complex and sensitive issues of national security. After today's ruling, if the political branches are intent on protecting the security interests of our nation by arresting and prosecuting those who would do the country harm, Congress and the President should also ensure that the United States Treasury is well-stocked to compensate the captured miscreants.

GOULD, Circuit Judge, dissenting:

I respectfully dissent because this case presents a nonjusticiable political question requiring scrutiny of an executive branch foreign policy decision. The majority passes judgment on an executive branch decision to act against foreign nationals on foreign soil in a matter directly affecting the United States's relations with a foreign nation.¹ Because our Constitution entrusts the conduct of our nation's foreign policy to Congress and to the executive branch, this case poses a political question decidedly inappropriate for judicial review.

1 Alvarez's apprehension by Mexican nationals was directed by the Drug Enforcement Administration (DEA), acting under authority delegated from the Attorney General.

The Supreme Court has held that certain allegations of illegal government action should not be ruled on by the federal courts even though all jurisdictional requirements are met. The Court has said that decision in these areas should be left to the politically accountable branches of government, the executive branch and Congress. The reasons for this political question doctrine have been variously described,² but, at its core, the political question doctrine represents the judiciary's wise recognition that the structure of our Constitution, its separation of powers, requires judicial restraint when certain subjects are at issue.

2 The political question doctrine minimizes judicial intrusion into the operations of the other branches of government. See Gilligan v. Morgan, 413 U.S. 1, 7, 93 S.Ct. 2440, 37 L.Ed.2d 407 (1973). It limits the courts' role in a democratic society. Alexander Bickel, The Least Dangerous Branch 184 (1962). It preserves the courts' legitimacy. Baker v. Carr, 369 U.S. 186, 267, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962) (Frankfurter, J., dissenting). And it allocates decisions to the branches of government that have superior expertise. See Fritz Scharpf, Judicial Review and the Political Question: A Functional Analysis, 75 Yale L.J. 517, 567 (1966).

The Supreme Court in Baker v. Carr made the classic statement of the considerations that animate the political question doctrine:

Prominent on the surface of any case held to involve a political question is found a textually demonstrable constitutional commitment of the issue to a coordinate political department; or a lack of judicially discoverable and manageable standards for resolving it; or an unusual need or the impossibility of deciding without an initial policy determination of a kind clearly for nonjudicial discretion; or the impossibility of a court's undertaking independent resolution without expressing lack of the respect due coordinate branches of government; for unquestioning adherence to a political decision already made; or the potentiality of embarrassment from multifarious pronouncements by various departments on one question.

369 U.S. 186, 217, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962). The Court held that any one of the above-listed characteristics may be sufficient to preclude judicial review. Id.

Most significantly for this case, the Court has held that the considerations described in Baker v. Carr bar judges from reviewing the executive branch's foreign policy decisions:

[T]he very nature of executive decisions as to foreign policy is political, not judicial. Such decisions are wholly confided by our Constitution to the political departments of the government, Executive and Legislative. They are delicate, complex, and involve large elements of prophecy. They are and should be undertaken only by those directly responsible to the people whose welfare they advance or imperil.

Chicago S. Airlines, Inc. v. Waterman S.S. Corp., 333 U.S. 103, 111, 68 S.Ct. 431, 92 L.Ed. 568 (1948). See also Oetjen v. Central Leather Co., 246 U.S. 297, 302, 38 S.Ct. 309, 62 L.Ed. 726 (1918) ("The conduct of the foreign relations of our Government is committed by the Constitution to the Executive and Legislat[ure,] 'the political' Departments of the Government, and the propriety of what may be done in the exercise of this political power is not subject to judicial inquiry or decision."); United States v. Curtiss-Wright Export Corp., 299 U.S. 304, 315, 57 S.Ct. 216, 81 L.Ed. 255 (1936) (holding nonjusticiable the question of the legality of a congressional resolution because "[t]he whole aim of the resolution [was] to affect a situation entirely external to the United States, and falling within the category of foreign affairs"). In other words, judges should not interfere in the management of our nation's foreign policy by opining (as the majority opines) on the executive's foreign policy decisions.

The DEA officials' decision to abduct Alvarez from Mexico without the Mexican government's permission was an executive foreign policy decision. Soon after a federal grand jury indicted Alvarez for murder, DEA officials negotiated with Mexican government officials for Alvarez's capture. When those negotiations proved unsuccessful, DEA officials decided it would be in the United States's interest to accomplish Alvarez's capture by means other than extradition. The officials' decision was consistent with a United States policy, formally announced in 1989, that it was permissible for law enforcement agencies such as the DEA to apprehend individuals accused of violating United States criminal law in foreign states without the consent of the foreign state. Barry E. Carter Phillip R. Trimble, International Law 794 (3d ed. 1999).

The foreign-policy nature of the DEA officials' decision to order Alvarez's capture in Mexico is amply demonstrated by the world's response to the incident. Mexico requested an official report on the role of the United States in the abduction and later sent diplomatic notes of protest from its embassy to the United States Department of State. Mexico also requested the arrest and extradition of the American law enforcement agents allegedly involved in the abduction. After the United States Supreme Court declined to invalidate Alvarez's capture, twenty-one member states of the Ibero-American Conference argued before the United Nations General Assembly that the extraterritorial exercise of criminal jurisdiction through the use of unilateral measures of coercion, such as abductions, violates the law of nations. Virginia Morris M.-Christiane Bourloyannis-Vrailas, The Work of the Sixth Committee at the Forty-Eighth Session of the UN General Assembly, 88 Am. J. Int'l L. 343, 357 (1994). The Iranian Parliament passed a draft law giving the president of Iran the right to arrest Americans who take action against Iranian citizens or property anywhere in the world and to bring them to Iran for trial. Carter and Trimble, supra, at 794. The Canadian Minster of External Affairs told the Canadian Parliament that any attempt by the United States to kidnap someone in Canada would be regarded as a criminal act and a violation of the U.S.-Canada extradition treaty. Id. at 792. The lower house of Uruguay's parliament voted that the United States's policy of unilaterally abducting foreign nationals showed "a lack of understanding of the most elemental norms of international law, and in particular an absolute perversion of the function of extradition treaties." Id. The presidents of Argentina, Bolivia, Brazil, Chile, Paraguay, and Uruguay issued a declaration expressing their concern. Id. These are only a few of the dozens of acts of diplomatic protest to the United States's policy of extraterritorial abduction, of which Alvarez's capture was the first well-publicized example.

The international response to Alvarez's capture was predictable, and executive officials may well have weighed it in determining whether to order that action. If the capture were merely an "exercise of the executive's law enforcement powers," as the majority suggests, rather than a foreign policy decision, the capture would not have become an international diplomatic incident. Indeed, the foreign-policy ramifications of international abduction have been well-known for decades, at least since the United Nations Security Council determined in 1960 that Israeli agents' abduction of the Nazi war criminal Adolf Eichmann from Argentina violated that country's sovereignty. Mary Alice Kovac, Apprehension of War Crimes Indictees: Should the United Nations' Courts Outsource Private Actors to Catch Them?, 51 Cath. U.L. Rev. 619, 631 (2002).

The foreign-policy nature of the decision to capture Alvarez is also demonstrated by the executive branch's policies governing this type of arrest. Executive branch officials have told Congress that agents follow

procedures designed to ensure that U.S. law enforcement activities overseas fully take into account foreign relations and international law. These procedures require that decisions as to extraordinary renditions from foreign territories be subject to full inter agency coordination and that they be considered at the highest levels of the government.

Carter and Trimble, supra, at 794. The Department of Justice has codified these special foreign-policy procedures in its United States Attorneys' Manual:

Due to the sensitivity of abducting defendants from a foreign country, prosecutors may not take steps to secure custody over persons outside the United States (by government agents or the use of private persons, like bounty hunters or private investigators) by means of Alvarez-Machain type renditions without advance approval by the Department of Justice. Prosecutors must notify the Office of International Affairs before they undertake any such operation.

Department of Justice, United States Attorneys' Manual, § 9-15.610 (1997). These executive branch procedures demonstrate that the capture of a foreign national on foreign soil is no ordinary law enforcement choice; rather, it is a serious foreign policy decision.

The DEA officials' decision to order the capture of a foreign national on foreign soil involved delicate foreign policy questions: Did the need to prosecute Alvarez for the torture and murder of an American official justify the United States's taking actions that might offend Mexico's government? Did 7316 the desire to send a tough message to foreign governments and drug cartels³ justify the United States's taking actions that might offend other nations' conceptions of international law? Did the benefits to the United States of ordering Alvarez's capture outweigh any potential diplomatic or trade retaliation Mexico might take?

3 Alan J. Kreczko, then Deputy Legal Adviser to the State Department, told Congress that some foreign governments "noted in private that [the United States policy of apprehending criminals overseas] will cause narcotics traffickers to have an increased fear of apprehension by the United States." Carter and Trimble, supra, at 794. Executive officials may have weighed this potential benefit to the United States when deciding whether to order Alvarez's capture on Mexican soil.

These questions, and other questions DEA officials may have considered (some of them beyond a judge's imagining), are beyond the competence and jurisdiction of federal judges. Unlike DEA officials, federal judges do not negotiate with foreign officials on behalf of the United States. Federal judges do not seek to influence foreign governments to accomplish the United States's policy objectives. Federal judges do not specialize in combating the international drug trade. Federal judges do not routinely work with the State Department. Federal judges do not commonly maintain relationships with overseas officials. Federal judges do not possess special training or knowledge relevant to drug interdiction on foreign soil.

Despite these deficits, the majority passes judgment on the executive branch's conduct of its constitutionally assigned responsibility. In doing so, the majority interferes with a delicate and complex decision "of a kind for which the Judiciary has neither aptitude, facilities nor responsibility." Waterman, 333 U.S. at 111, 68 S.Ct. 431. See also Baker, 369 U.S. at 211, 82 S.Ct. 691 (noting that foreign policy issues "frequently turn on standards that defy judicial application or involve the exercise of a discretion demonstrably committed to the executive or legislature").

The majority's decision injects a new and worrisome factor into executive officials' foreign policy calculus. Executive officials now must consider not only the international implications of their decisions, but also the possibility that federal judges, and eventually juries,⁴ will disapprove of their management of the nation's foreign policy. When executive officials consider whether to order action on foreign soil, they should properly weigh the possibility of diplomatic protests, international friction, and trade sanctions. They should not be required also to weigh the possibility of tort lawsuits by foreign nationals seeking punitive damages in their own nation's courts. They should not be required to consider the possible embarrassment such lawsuits will entail for them or for the United States. The majority's error allows the fear of tort lawsuits to taint officials' already complicated foreign policy analysis. Cf. Johnson v. Eisentrager, 339 U.S. 763, 789, 70 S.Ct. 936, 94 L.Ed. 1255 (1950) ("Certainly it is not the function of the Judiciary to entertain private litigation — even by a citizen — which challenges the legality, the wisdom, or the propriety of the Commander-in-Chief in sending our armed forces abroad or to any particular region.").

4 We should be concerned with the possibility that parties will urge juries with no foreign policy expertise to "send a message" to the United States government that particular foreign policy actions violate international norms of civilized conduct, as advanced by self-interested advocates. This is no way to conduct a foreign policy.

The majority's second-guessing of the executive branch's decision to capture Alvarez is also unseemly in that it prevents our government from speaking with a single voice on an important foreign policy decision. This feature of the majority decision is inconsistent with the Supreme Court's Baker v. Carr decision, which identified "multifarious pronouncements by various departments on one question" as an evil against which the political question doctrine guards. See Baker, 369 U.S. at 217, 82 S.Ct. 691. See also id. at 211, 82 S.Ct. 691 (noting that foreign policy questions "uniquely demand single-voiced statement of the Government's views"). By effectively taking the side of Mexico in this international dispute, the majority invades the province of the executive branch and embarrasses the United States on the international stage. The majority decision undermines not only the United States's foreign policy objectives, but also the unelected judiciary's fragile legitimacy with the American public.

The majority makes an unconvincing effort to limit its decision's scope.⁵ Despite that effort, the grave implications of the majority decision are plain. By holding that federal courts can review for compliance with international law executive branch decisions to act against foreign nationals on foreign soil, the majority transforms the executive branch's foreign policy decisions into occasions for judicial review. The detention and interrogation of suspected terrorists in Afghanistan; the bombing of Iraq; intelligence gathering around the world — under the majority's holding, these and other foreign policy actions may be the next subject of federal court litigation. The majority unintentionally licenses the United States's enemies, or those who desire to embarrass the United States, to use the federal courts for the purpose of propaganda or to accomplish more insidious ends. This use of the federal courts is altogether inappropriate, and the political question doctrine should guard against it.

5 Perhaps recognizing the manifold difficulties its decision creates, the majority attempts to limit its holding by reciting that it "does not speak to the authority of other enforcement agencies or the military, nor to the capacity of the Executive to detain terrorists or other fugitives under circumstances that may implicate our national security interests." I am concerned that the line the majority tries to draw in limiting its decision is more illusory than real.

The Supreme Court's past decisions in cases implicating foreign policy underscore the need to dismiss this case because it presents a nonjusticiable political question. The Supreme Court has held that it is improper for a court to review the executive's recognition of a foreign government, Oetjen, 246 U.S. at 302; to review the executive's determination that an individual claiming immunity possess diplomatic status, In re Baiz, 135 U.S. 403, 432, 10 S.Ct. 854, 34 L.Ed. 222 (1890); to review the executive's determination that a ship owned by a foreign nation is immune from in rem jurisdiction, Ex Parte Peru, 318 U.S. 578, 588-89, 63 S.Ct. 793, 87 L.Ed. 1014 (1943); to review the constitutionality of an undeclared war, Atlee v. Richardson, 411 U.S. 911, 911, 93 S.Ct. 1545, 36 L.Ed.2d 304 (1973) (affirming, without opinion, Atlee v. Laird, 347 F.Supp. 689, 705-07 (E.D.Penn. 1972)); to review the political branches' decision to prohibit arms sales to certain countries, Curtiss-Wright, 299 U.S. at 315, 329, 57 S.Ct. 216; to review the executive's assignment of foreign air routes, Waterman, 333 U.S. at 113-14, 68 S.Ct. 431; to review the political branches' determinations of when hostilities on foreign soil began or ended, Ludecke v. Watkins, 335 U.S. 160, 167-69, 68 S.Ct. 1429, 92 L.Ed. 1881 (1948); and, most relevant to this case, to judge whether the executive acted unlawfully in sending American troops to foreign soil. Eisentrager, 339 U.S. at 788-89, 70 S.Ct. 936.

An exercise of jurisdiction by federal courts in these cases deemed nonjusticiable by the Supreme Court — like the majority's exercise of jurisdiction in this case — would have divided our nation's government, pitting the judiciary against the political departments in matters involving our nation's conduct of foreign affairs. The majority's holding today that courts may review executive branch decisions to act against foreign nationals on foreign soil for compliance with international law poses at least as great a threat to our nation's foreign policy as did judicial review in the cases described above.⁶ The Supreme Court, for more than a century, has consistently rejected the attempts of interested litigants to use the federal courts as a weapon against the political branches' conduct of foreign affairs. It is unfortunate that the majority does not likewise recognize this restraint on our judicial power.

6 This case should have been dismissed not only as to the United States and the DEA agent defendant, but also as to Francisco Sosa, for he was acting under the direction of the United States, and it would be difficult to decide the propriety of his actions without commenting on the propriety of the executive's actions.

The Supreme Court's cases on the nonjusticiability of foreign policy decisions are not the only Supreme Court pronouncements that require us to reject Alvarez's lawsuit. The Supreme Court's enumeration in Baker v. Carr of the considerations relevant to determining whether a case presents a political question also requires dismissal. First, the majority's review of the executive branch decision to capture Alvarez ignores our Constitution's commitment of foreign policy decisions to Congress and the executive. See Baker, 369 U.S. at 217, 82 S.Ct. 691 (referring to "commitment of the issue to a coordinate political department" and a "policy determination of a kind clearly for nonjudicial discretion"). Second, it ignores the fact that federal judges lack the knowledge or training or tools to decide when such action is appropriate for the United States's foreign policy. See id. (referring to "a lack of judicially discoverable and manageable standards"). Third, it shows disrespect for the executive branch's management of foreign policy. See id. (referring to "the impossibility of a court's undertaking independent resolution without expressing lack of the respect due coordinate branches of government"). Fourth, it causes our government to speak with inconsistent voices about an issue on which our government should speak with one voice. See id. (referring to "the potentiality of embarrassment from multifarious pronouncements by various departments on one question" and "an unusual need for unquestioning adherence to a political decision already made").

Any one of the Baker v. Carr considerations is sufficient to render a case nonjusticiable. Id.; Made in the USA Found. v. United States, 242 F.3d 1300, 1312 (11th Cir. 2001). In this case, all of the Baker v. Carr considerations weigh sharply in favor of nonjusticiability. The majority errs by deciding the merits of this case.⁷

7 If I am mistaken that this case must be dismissed as presenting a non-justiciable political question, then I agree with Judge O'Scannlain's dissenting opinion that federal law authorized the DEA agents' conduct. See supra at 645 (en banc) (O'Scannlain, J., dissenting).

Human nature being what it is, and judicial nature following human nature, it is only natural that well-meaning judges will desire to comment on important affairs of the day involving political relations with other nations.⁸ But if the judiciary is to preserve its legitimacy, to show the respect due coordinate branches of government, and to avoid interfering in our nation's foreign relations, judges must show more restraint than the majority shows today. If ever there was a case in which the Supreme Court should revisit the political question doctrine and limit judicial interference in questions "in their nature political," Marbury v. Madison, 5 U.S. (1 Cranch) 137, 170, 2 L.Ed. 60 (1803), this is that case.⁹

8 It is not surprising that judges would want to promote the rule of law, including international law, particularly if they can have the last word on what that law requires. But that is no justification for an improvident exercise of an essentially political power concerning foreign affairs. The failure of my colleagues in the majority to address the political question doctrine, and the modest attention given it by my colleagues in dissent, see supra at 646-647 n. 2 (O'Scannlain, J., dissenting), might be explained by the government's failure to raise this issue. However, because the political question doctrine affects our jurisdiction, and because the political implications of this case for executive branch actions on foreign soil affecting relations with foreign nations are grave and apparent, I believe the issue must be considered sua sponte and fronted. Of course, it is not too late for the government, in seeking corrective review by petition for writ of certiorari, to advance the position that the political question doctrine precludes jurisdiction.

9 Even if this case did not present a nonjusticiable political question, it still should have been dismissed in part for lack of jurisdiction because the United States has not waived its sovereign immunity from lawsuits based upon the "discretionary functions" of government employees. See 28 U.S.C. § 2680(a). Because DEA officials' decision to order Alvarez's capture involved an element of judgment or choice based on considerations of public policy, United States v. Gaubert, 499 U.S. 315, 322-23, 111 S.Ct. 1267, 113 L.Ed.2d 335 (1991), it was a discretionary function that cannot be the subject of litigation against the United States. This jurisdictional flaw is not waivable, so the parties' failure to argue it does not justify our not considering it. See, e.g., In re Di Giorgio, 134 F.3d 971, 974 (9th Cir. 1998). The claims against the United States and its employees could not proceed. The United States's sovereign immunity would not, however, prevent our exercising jurisdiction over Alvarez's claims against Sosa.